Phd Thesis Emma Ghaziani, OT

UNIVERSITY OF COPENH AGEN FACULTY OF HEALTH AN D MEDICAL SCIENCES

PhD Thesis Emma Ghaziani, OT

Early therapeutic management of the affected arm functioning after stroke: prediction and intervention

Supervisors: S. Peter Magnusson, Christian Couppé, Hanne Christensen

This thesis has been submitted to the Graduate School of Health and Medical Sciences, University of Copenhagen, August 16, 2018

Name of department: Department of Physical and Occupational Therapy, Bispebjerg and Frederiksberg Hospital, Copenhagen, Denmark

Author: Emma Ghaziani, OT

Title and subtitle: Early therapeutic management of the affected arm functioning after stroke: prediction and intervention

Topic description: The present PhD-thesis focuses on the therapeutic management of the affected arm in early stroke rehabilitation. Specifically, the thesis examined the effectiveness of electrical somatosensory stimulation on the arm function and dexterity at 6 months post-stroke when the intervention was provided during the first 4 weeks after stroke. Furthermore, early (measured within 7 days post-stroke), easy-to- perform, clinical predictors of arm functioning at 6 months post-stroke were investigated.

Principal supervisor: S. Peter Magnusson, Professor, PT, DMSc Department of Physical and Occupational Therapy & Institute of Sports Medicine, Department of Orthopaedic Surgery M, Bispebjerg and Frederiksberg Hospital; Centre for Healthy Aging, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark

Co-supervisors: Christian Couppé, PT, PhD Department of Physical and Occupational Therapy & Institute of Sports Medicine, Department of Orthopaedic Surgery M, Bispebjerg and Frederiksberg Hospital; Centre for Healthy Aging, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark

Hanne Christensen, Professor, MD, DMSc Department of Neurology, Bispebjerg and Frederiksberg Hospital; Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark

Assessment Committee: Associate Professor Mette Aadahl (Chairperson) Department of Public Health, University of Copenhagen, Denmark

Associate Professor Iris Brunner Department of Clinical Medicine, Aarhus University, Denmark

Professor Marion Walker Faculty of Medicine & Health Sciences, University of Nottingham, United Kingdom

Submitted on: August 16, 2018 Grade: PhD Number of study units: 2

“If you can’t explain it simply, you don’t understand it well enough.” Albert Einstein

3 Table of contents

LIST OF SCIENTIFIC PAPERS ...... 7

LIST OF ABBREVIATIONS ...... 8

SUMMARY IN ENGLISH ...... 9

SUMMARY IN DANISH ...... 11

INTRODUCTION ...... 13

Stroke and stroke epidemiology ...... 13

The International Classification of Functioning, Disability and Health: a frame for describing consequences of stroke ...... 14

The effect of stroke on the arm functioning ...... 16

Time course of arm recovery after stroke ...... 16

Therapeutic interventions for recovery of arm functioning ...... 16 Electrical somatosensory stimulation ...... 18 Gap of knowledge ...... 19

Early prediction of arm functioning at 6 months post-stroke ...... 19 Gap of knowledge ...... 20

OBJECTIVES AND HYPOTHESIS ...... 20

METHODS ...... 21

Study design ...... 21 Paper I and II: Early intervention for the paretic arm after stroke ...... 21 Paper III: Early predictors of the affected arm functioning at 6 months post-stroke ...... 22

Study settings and participants ...... 23 Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) ...... 23 Paper III: Early predictors of the affected arm functioning at 6 months post-stroke ...... 26

Intervention ...... 28 Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) ...... 28

Assessment instruments ...... 30 4 Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) ...... 30 Paper III: Early predictors of the affected arm functioning at 6 months post-stroke ...... 34

Sample size ...... 35 Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) ...... 35 Paper III: Early predictors of the affected arm functioning at 6 months post-stroke ...... 35

Randomization ...... 35 Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) ...... 35

Blinding ...... 35 Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) ...... 35

Statistical methods ...... 36 Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) ...... 36 Paper III: Early predictors of the affected arm functioning at 6 months post-stroke ...... 36

KEY RESULTS ...... 37

Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) ...... 37

Paper III: Early predictors of the affected arm functioning at 6 months post-stroke ...... 53

DISCUSSION ...... 59

Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) ...... 59 Main results ...... 59 Comparison with studies showing beneficial effect of ESS ...... 59 Possible explanations for not finding a beneficial effect of ESS in the present trial...... 60 Strengths and limitations ...... 61

Paper III: Early predictors of the affected arm functioning at 6 months post-stroke ...... 63 Main results ...... 63 Comparison with other studies ...... 63 Strengths and limitations ...... 63

CONCLUSIONS ...... 64

FUTURE CONSIDERATIONS ...... 65

ACKNOWLEDGEMENTS ...... 66

REFERENCES ...... 67

5 APPENDIX (PAPER I, PAPER II, PAPER III) ...... 76

6 List of scientific papers

The present PhD-thesis is based on three scientific papers (I, II, III) focusing on two distinct topics related to early rehabilitation of arm functioning after stroke: intervention and prediction.

Topic 1: Early intervention for the affected arm functioning after stroke: I. Ghaziani E, Couppé C, Henkel C, Siersma V, Søndergaard M, Christensen H, Magnusson SP. Electrical somatosensory stimulation followed by motor training of the paretic upper limb in acute stroke: study protocol for a randomized controlled trial. Trials 2017; 18: 84.(1)

II. Ghaziani E, Couppé C, Siersma V, Søndergaard M, Christensen H, Magnusson SP. Electrical somatosensory stimulation in early rehabilitation of arm paresis after stroke: A randomized controlled trial. Neurorehabilitation & Neural Repair (provisionally accepted)

Topic 2: Early predictors of the affected arm functioning at 6 months post-stroke: III. Ghaziani E, Couppé C, Siersma V, Christensen H, Magnusson SP, Sunnerhagen KS, Persson HC, Alt Murphy M. Can clinical tests early post-stroke aid the prediction of arm functioning at 6 months? (in preparation)

7 List of abbreviations

ADL Activities of daily living BBT Box and Block Test BFH Bispebjerg and Frederiksberg Hospital, Copenhagen, Denmark DALYs Disability-adjusted life years = years lost due to premature death + years of healthy life lost due to disability; measure of disease burden ESS Electrical somatosensory stimulation ESS-trial The intervention study presented in Paper I and II FMA-UE Upper extremity section of Fugl-Meyer Assessment (subscale A−J, 0−126 points) FMA-UE-AD Upper extremity section of Fugl-Meyer Assessment, subscale A−D (0−66 points) GBD Study Global Burden of Disease Study HS Haemorrhagic stroke ICC Intra-class correlation coefficient IS Ischaemic stroke MCID Minimally clinically important difference MDC Minimal detectable chage mRS modified Rankin Scale OR Odds ratio OT(s) Occupational therapy/Occupational therapist(s) PT(s) Physical therapy/Physical therapist(s) PTT Perceptual threshold of touch SALGOT-study the Stroke Arm Longitudinal Study at the University of Gothenburg SGPALS Saltin-Grimby Physical Activity Level Scale TENS Transcutaneous electric nerve stimulation WHO World Health Organization

8 Summary in English

Background: Arm paresis is present in 48–77% of acute stroke patients. Complete functional recovery is reported in only 12–34%. Although the arm recovery is most pronounced during the first 4 weeks post-stroke, few studies examined the effect of upper extremity interventions during this period. To be applicable in clinical practice, prognosis of arm recovery needs to be based on easy-to-perform, meaningful measures. Several clinical tests have been proposed for prognosis of arm functioning after stroke; further validation of their predictive value is needed (paper I, II, III).

Objectives: a) To investigate the effect of electrical somatosensory stimulation (ESS) delivered during early stroke rehabilitation on the recovery of arm functioning; b) to examine the individual predictive value of easy-to-perform clinical tests for early prognosis of arm functioning (paper I, II, III).

Methods: 102 patients with arm paresis were randomized to an intervention or a control group within 7 days post-stroke according to our sample-size estimation. The intervention group received 1-hour ESS to the paretic arm daily during hospitalization immediately followed by minimum 15-minute task-oriented arm training that was considered a component of the usual rehabilitation. The control group received a placebo ESS followed by identical training. Primary outcome: Box and Block Test (BBT); secondary outcomes: upper extremity section of Fugl- Meyer Assessment, subscale A−D (FMA-UE-AD), grip strength, pinch strength, perceptual threshold of touch, pain and modified Rankin Scale (mRS); all recorded at baseline, post- intervention and at 6 months post-stroke. Furthermore, the following variables measured 3−7 days post-stroke using FMA-UE were considered potential predictors of FMA-UE-AD at 6 months post-stroke: shoulder abduction and elbow extension within synergies, forearm pronation/supination, wrist dorsiflexion, mass finger extension, grasping ability, and sensory function. Based on merged data from two independent studies (n=223), logistic regression was used for each predictor to calculate the odds ratio of two levels of arm functioning: FMA-UE- AD ≥32 and ≥58 (paper I, II, III).

Results: There were no differences between the intervention and the control groups for any outcome measures at any time points. Clinically significant improvements were observed for FMA-UE-AD, hand grip strength and mRS in both groups. Moreover, patients with initial partial 9 shoulder abduction were at least 7.3 times more likely to achieve a FMA-UE-AD≥32 at 6 months post-stroke. The probability of a FMA-UE-AD≥58 was at least 3.3−35.2 times higher in patients with partial/full distal movement (forearm pronation/supination, wrist dorsiflexion and grasping ability) compared with patients with absent movement. Patients with full elbow and mass finger extension were at least 36.8 and 18.3, respectively, times more likely to achieve a FMA-UE-AD≥58 than those with no movement. Full sensory function had a significant but more modest predictive value (paper II, III).

Conclusions: The data show that the present ESS-protocol prior to arm training was equally beneficial as arm training alone. These results apply for patients with mild-to-moderate stroke and moderate arm impairments. It cannot be excluded that patients with other characteristics, during other time intervals post-stroke or using a different ESS-protocol might benefit. Furthermore, this thesis confirmed that sufficient sensory function and some proximal/distal arm movement early post-stroke predict a better arm functioning at 6 months in patients with same characteristics; partial/full distal movement was identified as a predictor of a FMA-UE-AD≥58 at 6 months post-stroke (paper II, III).

10 Summary in Danish

Baggrund: 48−77% af alle apopleksi patienter oplever armparese i den akutte sygdomsfase, hvoraf kun 12−34% remitterer fuldstændigt. Dette er problematisk, da en næsten fuld funktion er nødvendig for, at patienterne efterfølgende kan involvere den afficerede arm i udførelsen af daglige aktiviteter. Til trods for at størstedelen af armens funktionsevne remitterer inden for de første 4 uger efter apopleksi, har kun få studier undersøgt effekten af rehabiliteringsinterventioner rettet mod den paretiske arm i dette tidsinterval. Af hensyn til implementeringen i klinisk praksis bør vurderingen af den paretiske arm foretages ved hjælp af undersøgelsesredskaber, som er let tilgængelige, nemme at anvende, meningsfulde for patient/kliniker samt prædikterer fremtidig funktionsevne. Flere kliniske tests har været foreslået og yderligere validering af deres prædiktive værdi er nødvendig.

Formål: a) At undersøge effekten af elektrisk somatosensorisk stimulation (ESS) på remissionen af armens funktionsevne, når interventionen bliver igangsat tidligt efter apopleksi, b) At undersøge den individuelle brugbarhed af kliniske tests, som nemt kan foretages i den akutte apopleksi fase, til prædiktion af armens fremtidige funktionsevne.

Metode: 102 patienter med armparese af forskellige sværhedsgrader blev randomiseret til kontrol- eller interventionsgruppe i løbet af den første uge efter apopleksi. Deltagerne i interventionsgruppen fik 1 times ESS efterfulgt af minimum 15 minutters opgaveorienteret armtræning dagligt under deres hospitalsindlæggelse. Armtræningen var en del af standardrehabiliteringen på hospitalet. Kontrolgruppen modtog en placebo ESS efterfulgt af samme type armtræning og øvrig standardrehabilitering. Den primære effektmål var Box and Block Test (BBT). De sekundære effektmål var: armsektionen af Fugl-Meyer Assessment, subscale A−D (FMA-UE-AD), håndgrebsstyrke, fingergrebsstyrke, sensorisk funktion, smerte og modified Ranking Scale (mRS). Data for alle effektmål blev indsamlet ved baseline, endt intervention og 6 måneder efter apopleksi. Brugbarheden af skulder abduktion og albue ekstension i synergier, pronation/supination af underarm, dorsifleksion af håndled, finger masseekstension, grebsfunktion og sensorisk funktion blev undersøgt for prædiktion af armens funktionsevne 6 måneder efter apopleksi. De potentielle prædiktorer blev målt 3−7 dage efter apopleksi ved hjælp af FMA-UE. Vi anvendte logistisk regression til at beregne odds ratio for at opnå en funktionsevne svarende til FMA-UE-AD ≥32 og ≥58 for hver potentiel prædiktor. Beregningen blev baseret på et fusioneret datasæt fra to uafhængige studier.

Resultater: Vi fandt ingen forskel mellem interventions- og kontrolgruppen for ingen af effektmålene og på intet tidspunkt. Til gengæld fandt vi, at FMA-UE-AD, håndgrebsstyrke og mRS udviklede sig positivt og klinisk signifikant i begge grupper. Sandsynligheden for at patienter med delvis skulder abduktion i løbet af den første uge efter apopleksi ville opnå FMA-UE-AD≥32 6 måneder efter apopleksi var mindst 7.3 gange højere end hos patienter uden skulder abduktion. Sandsynligheden for FMA-UE-AD≥58 var mindst 3.3−35.2 gange højere hos patienter med delvis/fuld distal bevægelighed (pronation/supination af underarm, dorsifleksion af håndled og grebsfunktion) end hos patienter uden aktiv distal bevægelighed i armen. For patienter med fuld albue og finger massekstension var sandsynligheden for FMA-UE-AD≥58 hhv. 36.8 og 18.3 gange højere end i patientgruppen uden aktiv bevægelighed. Fuld sensorisk funktion havde en signifikant, men mere beskeden prædiktiv værdi.

Konklusion: Den nuværende ESS-protokol efterfulgt af armtræning er lige så effektiv som armtræning alene. Dette resultat er gældende for patienter med mild til moderat apopleksi og moderat funktionsnedsættelse i armen. Vi kan ikke udelukke muligheden for, at ESS kan have gavnlig effekt når det anvendes på patienter med andre karakteristika, i andre tidsintervaller efter apopleksi eller i form af en anden protokol. Denne afhandling bekræfter tidligere fund om, at sufficient sensorisk funktion og proximal/distal bevægelighed i armen tidligt efter apopleksi er associerede med en bedre arm funktionsevne 6 måneder senere hos patienter med ovenstående karakteristika. Endvidere viser vores resultater, at tidlig partiel/fuld distal bevægelighed i armen prædikterer FMA-UE-AD≥58 6 måneder efter apopleksi.

12 Introduction

Stroke and stroke epidemiology Stroke is defined by the World Health Organization (WHO) as a cerebrovascular disease caused by the interruption of the blood supply to the brain, usually because a blood vessel bursts (haemorrhagic stroke, HS) or is blocked by a clot (ischaemic stroke, IS). This cuts off the flow of oxygen and nutrients, damaging the brain tissue (2). Depending on the stroke location and the size of the affected brain areas, patients experience various degrees of impaired body functions, difficulties in managing activities of daily living (ADL) and restrictions in their social lives (3- 5). Factors such as hypertension, current smoking, abdominal obesity, unhealthy diet (e.g. increased consumption of meat, eggs, salty snacks, fast food, cooking with lard), sedentary life, diabetes mellitus, alcohol intake, depression, cardiac causes (e.g. atrial fibrillation, previous myocardial infarct), and ratio of apolipoproteins B to A1 have shown to increase the risk of developing the disease (6).

The Global Burden of Disease (GBD) Study (7) was initiated in 1992 at the request of the World Bank and conducted in collaboration to WHO. The objective of the GBD Study was to perform worldwide, longitudinal, systematic epidemiological assessments for major diseases to inform health care policies development and to support disease prevention and control. According to the GBD 2013 Study (8, 9), stroke was the second largest cause of death from all causes (11.8% [95% CI: (10.9−13.0%)]) and the second largest cause of burden of disease from all causes measured by disability-adjusted life years (DALYs= years lost due to premature death + years of healthy life lost due to disability) (4.5% [95% CI: (4.1−5.2)]). The percentage contribution of stroke to the total burden of disease was not statistically different from the contribution of the top leading cause, ischemic heart disease (6.1% [95% CI: (5.5−6.8)]), especially in the developed countries. Despite a significant drop in age-adjusted DAYLs and mortality rates of IS and HS combined per 100,000 people from 2,431 (95% CI: [2,224−2,631]) and 142 (95% CI: [129−14]), respectively, in 1990 to 1,807 (95% CI: [1,667−1,992]) and 110 (95% CI: [102−122), respectively, in 2013, the absolute number of new stroke events, stroke survivors, and deaths increased significantly for both IS and HS. Notably, the global absolute number of stroke survivors almost doubled between 1990 and 2013 for both IS and HS. A statistically significant increase in the absolute number of DAYLs occurred only for IS in the same period. Thus, in 2013 there were approximately 25.7 million stroke survivors (71% with IS), 10.3 million new

13 strokes (76% caused by IS), 6.5 million deaths from stroke (51% died from IS), and 113 million DAYLs (42% due to IS) worldwide. The increase in the absolute number of stroke survivors was presumably a result of demographic changes (ageing and growing population), improved stroke care, improved diagnostic methods (imaging techniques) that enables the detection of minor strokes, insufficient reduction of behavioural risk factors, and an increased presence of stroke in the younger population. Between 1990 and 2013, significant increases in the absolute numbers of prevalent cases, deaths and DALYs due to IS and HS were reported among younger adults aged 20−64 years (10). Stroke in working-age population has massive personal and societal consequences due to a larger number of DAYLs and an impaired ability to manage responsibilities as wage earner and caregiver in the family. This development was likely attributable to unfavourable changes in behavioural risk factors among the younger population such as unhealthy diets high in sugar, salt and processed foods, smoking, alcohol intake, drug use and reduced level of physical activity (11, 12). It has been suggested that stroke should no longer be considered a disease of the elderly, as 2/3 of all strokes currently occur among people < 70 years of age (8).

The Danish Health Authority estimates that 93,000 stroke survivors are currently living in Denmark, and the number of incident strokes is 15,000 per year. Yearly, there are 3,600 reported deaths due to stroke, corresponding to 7% of deaths by all causes. The cost of home care services for stroke survivors is the highest among all diseases, stroke representing the top leading cause of disability in Danish adults (13) .

The International Classification of Functioning, Disability and Health: a frame for describing consequences of stroke The International Classification of Functioning, Disability and Health (ICF) (14) is a widely accepted conceptual model and classification of human functioning and disability. The model and the terminology proposed by ICF will be used in this PhD-thesis to describe the multifaceted effect of a stroke (ICF: health condition) on individuals. The concept of body functions describes the physiological functions of body systems. Impairments are dysfunctional body functions (e.g. decreased hand grip and pinch strength, sensory dysfunction, impaired ability to perform volitional arm movements). Activity is the execution of an action (i.e. a motor skill) or a task (i.e. an ADL), and activity limitations are difficulties in executing these activities (e.g. difficulties in manipulating objects or performing ADLs). Within the concept of activity, ICF makes a distinction between capacity/ability and actual performance. Capacity indicates an 14 individual’s highest level of ability to perform a task or an action and is typically assessed in standardized environments. The majority of the arm tests used in stroke rehabilitations represents measures of capacity. Actual performance, on the contrary, is a measure of what the individual does in the real-life environment; in recent years, this aspect of activity have been successfully assessed using accelerometers (15). Participation denotes involvement in life situations; participation restrictions are problems an individual may experience in involvement in life situations (e.g. not being able to fulfil family or occupational roles). Furthermore, the ICF operates with environmental and personal factors. Because the domains of participation, environmental and personal factors do not represent key concepts in the present PhD-thesis, they will not be presented in more details. Human functioning is an umbrella term, consisting of body functions, activities and participation. Conversely, impaired human functioning is labelled disability and encompasses impairments, activity limitations and participation restrictions. Figure 1 visualizes the interaction between the components of the ICF, and how these components are covered the in present PhD-thesis (in italic).

Fig. 1: The ICF model of human functioning and disability adapted for use in the present PhD- thesis

Abbreviations: ADL: activities of daily living; BBT: Box and Block Test; FMA-UE: upper extremity section of Fugl-Meyer Assessment; PTT: perceptual threshold of touch; mRS: modified Rankin Scale

15 The effect of stroke on the arm functioning Arm paresis is a highly prevalent impairment after stroke. Paresis is a deficit in both strength and motor control (i.e. the ability to make coordinated, accurate and goal-directed movements) (16). It has been shown that arm paresis causes limitations in performing ADLs and may influence the quality of life detrimentally (17). Longitudinal cohort studies reported that arm paresis affects approximately 48–77% of patients at stroke onset (18-21). However, despite clinically important improvements (i.e. ≥ minimally clinically important difference, MCID) in arm function and functional capacity during the first 12 months post-stroke, the actual use of the affected arm in daily life does not seem to improve significantly (15). Evidence suggests that the affected arm will not routinely be used for daily tasks unless an almost complete recovery is acquired (22). Hence, there seems to be a disparity between the individuals’ capacity for activity performance and their actual real-arm-use in daily life. Despite rehabilitation, only 12–34% of stroke patients will achieve full arm recovery at 6 months post-stroke (23-25), most of the patients remaining with a non-functional arm in the long-term. The daily use of the affected arm at 12 months after stroke was estimated to be around 35% of the daily use of the non-affected arm (15), which, in fact, also seems to be affected by the stroke. It has been showed, that the actual activity performance with the non-paretic arm is lower in adults with chronic stroke than in adults without stroke (26).

Time course of arm recovery after stroke The concept of recovery usually refers to improved performance without distinguishing between true neurological recovery, which reflects the process of restitution of body functions, and the use compensatory strategies (27, 28). The greatest degree of arm recovery has been reported rapidly during the first 4 weeks after stroke (29-33), presumably due to spontaneous restitution of body functions. The first 4 weeks post-stroke seem to be of particular relevance for the rehabilitation process, as hand dexterity (i.e. motor skills as reaching, grasping, moving and releasing objects) also appears to be determined during this critical period (24). The recovery of arm functioning continues until 3−6 months after stroke, but to a lesser extent (19, 30, 33). Beyond the first 6 months, further improvements can be expected in 5–10% and deterioration in 15−25% of stroke patients (30).

Therapeutic interventions for recovery of arm functioning Several interventions are currently used by physiotherapists (PTs) and occupational therapists (OTs) to aid in the recovery of arm functioning after stroke. Although many studies have 16 investigated the effectiveness of various therapeutic interventions, the current evidence is still scarce or of modest quality. There is limited evidence that hands-on therapy (e.g. passive joint mobilization, manual stretching of soft tissue and passive exercises) is effective (34). Strength training in chronic stroke may reduce motor impairments in patients with mild-to-moderate arm paresis, but without any proven transfer into improved ADL-performance (35). The evidence for using orthoses and others supporting devices is inconclusive (36-38). A recent meta-analysis showed that mirror therapy is likely to improved arm function immediately after intervention, but the long-term effect is unclear (39). Another recent systematic review (40) concluded based on limited data that therapist-delivered sensory stimulation does not seem to facilitate the reduction of arm impairments or the improvement of ADL-performance. There is low-to moderate evidence that repetitive task-oriented practice improves arm function and motor skills; these improvements seem to be sustained up to 6 months post-intervention (41). Whether repetitive task-oriented practice can influence ADL-performance beneficially and whether the effect can be maintained beyond the first 6 months after ended intervention are still questions to be answered. When delivered as electromechanical and robot-assisted training, there is some evidence that repetitive task-oriented practice may help in improving arm function (including strength) and ADL-performance (42), but the evidence is currently of low-to-very-low quality. Compared to electromechanical and robot-assisted training, the use of virtually reality systems can provide motivating training environments with enhanced possibilities of training intensity and sensory feedback, which have shown to facilitate motor relearning. Incorporating virtual reality systems in arm rehabilitation seems to improve arm function (43); future studies remain to investigate their effectiveness on motor skills and ADL-performance. Constrained-induced movement therapy (CIMT) has received a lot of attention from researchers during the last decades. Despite initial promising results, the most recent Cochrane review (44) fails to demonstrate a superiority of various protocols for CIMT in reducing disability compared with other therapeutic interventions or no intervention. Though, CIMT is likely to improve arm motor function and motor skills. The long-term effect of CIMT is currently inconclusive. Electrical stimulation (ES) is another extensively investigated intervention. ES can induce a muscle contraction, or it can be a somatosensory stimulation below the motor threshold (45). Regardless the type of stimulation, there is some evidence that ES can aid in reducing motor impairments, but the questions regarding the optimal stimulation protocol (e.g. current intensity, pulse frequency, placement of electrodes, treatment frequency and duration), long-term effect and transfer of training effect into ADL-performance remain unanswered (45, 46). As this body of evidence is primarily based on studies conducted on ES in chronic stroke patients, it also remains 17 unknown to what extent ES applied in the acute phase after stroke affects the recovery of arm functioning.

Although most rehabilitation is delivered within the first 4 weeks after stroke onset, less than 10% of motor rehabilitation trials are initiated during this period of time (47). Regarding the first 4 weeks post-stroke, a recent systematic review and meta-analysis (48) concluded that there is sufficient level of evidence to recommend the use of modified CIMT and task-oriented training in routine clinical practice; the use of biofeedback and ES with motor response is recommended only as supplementary therapies based on the available evidence.

Finally, the optimal timing (i.e. acute, subacute or chronic stroke phase), intensity of practice (i.e. frequency, duration) and patient characteristics (i.e. stroke severity, severity of arm paresis, cognitive abilities) are important issues to consider and are still unsolved. Interestingly, finding the best intervention for arm recovery have been identified by stroke survivors, carers and health professionals as one of the top ten research priorities relating to life after stroke (49).

Electrical somatosensory stimulation It is widely accepted that somatosensory inputs are essential for (re)learning of motor skills and for motor performance, and stroke patients with intact somatosensory function are more likely to experience a better response to rehabilitation (50, 51). It has been suggested that augmenting somatosensory input to the affected arm after stroke has the potential to increase the excitability of the corresponding cortical area, leading to long-term plastic changes and improved motor recovery (52). Alternatively, it has been proposed that the interhemispheric balance, according to which the two hemispheres exert a mutual inhibition on each other in the healthy brain, may be disrupted by stroke. The restoration of this balance, either by increasing the cortical excitability of the affected hemisphere or by reducing the excitability of the unaffected one, might be beneficial for arm recovery (53). In healthy persons, low-intensity ES with no or weak motor response administrated to several locations on the arm (peripheral hand nerves, forearm muscles, the whole hand) has shown to increase the cortical excitability (54-60), and the effect outlasted the stimulation period (55, 56, 59). For example, a 2-hour suprasensory electrical somatosensory stimulation (ESS) to the peripheral hand nerves at the wrist has shown to increase cortical excitability in healthy persons (56, 60), while 30-minute suprasensory ESS of the whole hand was sufficient to the increase cortical excitability; the effect sustained for at least 1 hour after

18 ended ESS (54, 55). Suprasensory ESS was defined as the highest current amplitude that elicits paresthesia without any discomfort, pain or visible muscle twitches. The investigation of the ESS’s benefit in the rehabilitation of the paretic arm after stroke is still in its early stage. Studies showed that a single 2-hour session of ESS administrated to the peripheral hand nerves improves pinch strength, movement kinematics and motor skills in all stroke phases (61-66); of these studies, only one used ESS in addition to motor training (61). Furthermore, the application of multiple ESS-sessions to the peripheral hand nerves seems to improve arm motor skills in subacute (67) and chronic stroke (68) when followed by motor training. Though, the current evidence is scarce and conflicting (69, 70). Importantly, ESS is passive in nature, causes patients minimal discomfort, has no adverse effects, can easily be incorporated in regular practice, and the electrical device is relatively inexpensive (71).

Gap of knowledge In the planning phase of this PhD-project, the effectiveness of multiple ESS-sessions used in conjunction with motor training in early arm rehabilitation after stroke was not yet investigated in a randomized controlled trial (RCT) design. We considered that conducting a RCT would be valuable for stroke research and clinical practice of several reasons: a) There was limited evidence on which therapeutic arm interventions are effective when administrated early after stroke; b) ESS has a broad clinical applicability. ESS seemed to be suitable for patients with various degrees of arm paresis and cognitive function, patients with severe arm paresis and cognitive dysfunctions being the ones most in need for effective rehabilitation options. c) ESS could easily be implemented in clinical practice due to a low cost, and easy-to- deliver protocols.

Early prediction of arm functioning at 6 months post-stroke It is of particular relevance for clinicians to make accurate, early predictions of the long-term arm recovery using bedside tests that are easy-to-perform (i.e. are time-effective and do not rely on inaccessible technologies/expertise to the most hospital departments), and suitable for the majority of the patients regardless their cognitive status and severity of arm paresis. Firstly, this prediction is useful for determining the optimal proportion of restorative and compensatory strategies in the treatment plan. Secondly, the prediction is needed when informing the patients and their families

19 about the expected recovery potential, allowing them to adjust to the new life situation and plan for the future.

Several measures easily to perform during the clinical bedside examination in the acute/early subacute stroke phase have been proposed for the prediction of short- and long-term recovery of arm functioning: finger extension (25, 72, 73), shoulder abduction (25, 73), shoulder and elbow control (27), muscle strength (23, 74-76), and sensory function (75-77). According to a systematic review with meta-analysis (78) there is strong evidence that higher arm function and functional capacity, and the presence of motor and somatosensory evoked potentials at baseline are associated with better arm recovery. Furthermore, the review showed moderate evidence that higher leg function and less global disability are predictors for better arm recovery. The neurophysiological assessments require, though, special equipment and expertise, and they are consequently not employed in clinical practice on a regular basis. In line with recommendations from other reviews (79, 80), the authors emphasized the need for additional studies to further the development and validation of clinical predictors that easily can be used in regular clinical practice.

Gap of knowledge Further high-quality cohort studies are needed to validate clinical tests that can be easily performed in early therapeutic examinations and used to make predictions of future levels of arm recovery after stroke.

Objectives and hypothesis

The objective of this PhD-thesis was to enhance the knowledge of the therapeutic management of the paretic arm in early stroke rehabilitation by: a) investigating whether ESS delivered during the first 4 weeks post-stroke is more effective than usual rehabilitation alone in improving the arm functioning at 6 months post-stroke, b) examining the prediction value of easily performed, clinical, bedside tests for the early prognosis of arm functioning at 6 months post-stroke.

Specifically, we addressed the following research questions: a) Does ESS improve motor and sensory arm function, hand dexterity, and reduce global disability at the end of the intervention period?

20 b) Are possible benefits observed at the end of the ESS-intervention period still present or improved at 6 months post-stroke? c) What is the individual value of shoulder abduction and elbow extension within synergies, forearm pronation/supination, wrist dorsiflexion, mass finger extension, pincer grasp, cylinder grasp, and sensory function assessed within 7 days post-stroke for the prognosis of arm functioning at 6 months post-stroke?

Our hypotheses were: a) ESS followed by arm training during early hospitalization after stroke is superior to arm training alone regarding the arm dexterity measured by the Box and Block Test (BBT, primary outcome measure) at 6 months post stroke, and b) improvements in dexterity would be accompanied be reductions in impairments and global disability. Furthermore, we expected that our data would validate previously proposed clinical tests such as shoulder abduction, finger extension and sensory function as predictors of arm functioning measured with FMA-UE-AD at 6 months after stroke; the analyses for the remaining potential predictors were exploratory.

Methods

Study design

Paper I and II: Early intervention for the paretic arm after stroke Paper I and II reported a randomized single-blinded controlled trial (RCT) with two parallel arms and blinded endpoint adjudication. The intervention consisted of ESS to the paretic arm immediately followed by arm training. ESS was provided in addition to usual rehabilitation; the arm training was considered a part of the usual rehabilitation. ESS/placebo ESS was initiated within the first week post-stroke and provided no longer than 4 weeks post-stroke. In the remaining part of the present PhD-thesis, this trial will be mentioned as the ESS-trial. Figure 2 depicts the key steps in the ESS-trial.

21 Figure 2: Key steps in the ESS-trial

Paper III: Early predictors of the affected arm functioning at 6 months post-stroke Paper III reported a prediction study based on a prospective, longitudinal design. This study is a secondary analysis of merged data from two independent cohorts of stroke patients. The first cohort was derived from the Stroke Arm Longitudinal Study at the University of Gothenburg, Sweden (the SALGOT-study), with the aim of describing the recovery of arm functioning during the first 12 months after acute stroke (81, 82). The second cohort was the sample of participants from the ESS- trial presented in Paper I and II. The intervention and the control group in the ESS-trial could be merged because the trial demonstrated no difference between the effect of active ESS and placebo ESS (paper II, in press). Figure 3 shows the key steps in the prediction study.

22 Figure 3: Key steps in the prediction study

Study settings and participants

Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) The trial participants were recruited at the stroke rehabilitation unit of Bispebjerg and Frederiksberg Hospital (BFH), Copenhagen, Denmark. After a few days in the acute stroke unit, patients are usually transferred to the stroke rehabilitation unit that offers inpatient rehabilitation services while patients are still in need for constant medical care. These rehabilitation services are covered by the Danish national health insurance. The recruitment process occurred in several steps as illustrated in Figure 4: • Step 1: All patients consecutively admitted at the stroke rehabilitation unit and diagnosed with acute stroke (IS or HS) confirmed by magnetic resonance imaging (MRI) or computer tomography (CT) scan were screened for a newly developed arm paresis from 13 October 2014 to 1 March 2017, except a total of 6 months (holidays, recruitment and training of new trial staff). In total, 1214 stroke patients were identified; of these, at least 537 had a newly developed arm paresis. The number of patients with a new arm paresis was actually higher, as the screening log from the initial 5 months was incomplete.

23 • Phase 2: The stroke patients with a new arm paresis were subsequently screened for following eligibility criteria: o Age ≥ 18 years; o Residence in the hospital’s catchment area for stroke rehabilitation; o Complete recovery of the affected arm from a previous stroke; o Arm paresis as indicated by a score < 66 on the upper extremity section of the Fugl- Meyer Assessment, subscale A−D (FMA-UE-AD); o Possibility to initiate the ESS-intervention within 7 days post-stroke due to medical and logistical reasons; o No contraindications to ESS (e.g. pacemaker, skin impairment) (83); o Absence of cognitive dysfunctions or poor communication skills in Danish that limited the ability to provide informed consent; o No severe pre-stroke disability as indicated by a modified Rankin Scale (mRS) score of 5; o The patient did not participate in other biomedical intervention trials within the last 3 months. At least 413 patients (incomplete screening log) did not meet all the eligibility criteria. • Phase 3: Eligible patients were asked to provide written, informed consent before trial enrolment. More than 8 patients declined to participate, and at least 14 were excluded of other reasons (incomplete screening log). • Phase 4: A number of 120 patients were enrolled in the trial; 49 were randomized to the control group and 53 to the intervention group.

24 Figure 4: The flow of participants through the ESS-trial

25 Paper III: Early predictors of the affected arm functioning at 6 months post-stroke The participants in the prediction study comprised the merged SALGOT-ESS-cohort. The ESS- cohort has been presented in the above-mentioned section of this PhD-thesis. In the SALGOT- study (81, 82), all patients consecutively admitted at the largest of the three stroke units at the Sahlgrenska University Hospital, Gothenburg, from February 2009 to December 2010 were screened for following inclusion criteria: o first-ever acute stroke (IS or HS); o impaired arm function at day 3 after stroke as indicated by FMA-UE-AD <66; o admission to the stroke unit within 3 days after stroke onset; o residence in the Gothenburg urban area. Patients were excluded if one of the following criteria was present: o injury/condition prior to the stroke that limited the use of the affected arm; o severe, multiple impairments or diminished physical condition prior to stroke; o short life expectancy; o non-Swedish speaker. The merged SALGOT-ESS-cohort comprised 223 participants; 102 from the ESS-trial and 121 from the SALGOT-study. The inclusion process in the prediction study is shown in Figure 5.

) =66

(n=1 AD

ed to the stroke stroke the ed to

stroke)

stroke (n=335) (n=335) stroke

: FMA

)

(n=641) (n=641)

(n=122

(3 days post (3 days

)

EXCLUDED

stroke (n=10) stroke

)

SAMPLE

STROKE EXCLUDED EXCLUDED

assessment + refuse to participate (n=1) participate to + refuse assessment

from February 2009 to December 2010 2010 December to 2009 February from

POST SALGOT

admitt consecutively patients All

Swedish speakers (n=8) speakers Swedish

up (recurrent stroke, arm fracture) (n=11) fracture) arm stroke, (recurrent up

DAYS DAYS

(

Missed for inclusion (n=5) inclusion Missed for

Decline to participate (n=36) participate to Decline

Missed for screening (n=43) screening Missed for

Non

Discharged < 72h post < 72h Discharged

Severe multiple impairments (n=90) impairments multiple Severe

Prior arm impairment (n=58) impairment arm Prior

Residence outside the hospital catchment area (n=56) area catchment hospital the outside Residence

No arm impairment at day 1−2 post 1−2 day at impairment No arm

OF CANDIDATE PREDICTORS AND CONFOUNDERS CONFOUNDERS AND PREDICTORS CANDIDATE OF

•

ELIGIBILITY: ELIGIBILITY: for eligibility (n=763) eligibility for

Moved (n=3) Moved

Not possible to contact (n=3) contact to Not possible

Not performed baseline assessment + dead (n=1) + dead assessment baseline Not performed

Dead (n=15)

Not followed Refused to participate (n=13 participate Refused to

baseline Not performed

•

STROKE) (n=176) STROKE)

- ASSESSEMENT ASSESSEMENT

Not performed (n=2) Not performed

or HS with IS diagnosed t and

were screened were

uni

ASSESSED FOR FOR ASSESSED

•

SAMPLE (n=223) SAMPLE -

POST MONTHS (6

for

ESS

AD -

t and t and SALGOT

FMA

were screened were ) (Screening log (Screening )

holidays,

hs hs

214)

413)

(n=102)

March 2017 March

ASSESSMENT OF ASSESSMENT

)

(n=

) (n=1,

a total of 6 months ( months 6 of total a

STROKE

staff

criteria criteria

er limb paresis (n=537 paresis er limb

SAMPLE

for the initial 5 mont 5 the initial for

ESS

DAYS POST DAYS

incomplete)

ASSESSED FOR ELIGIBILITY 1 ELIGIBILITY FOR ASSESSED

incomplete)

ASSESSED FOR ELIGIBILITY 2 ELIGIBILITY FOR ASSESSED

from October 2014 to to 2014 October from

paresis, except except paresis,

≤ 7 ≤

(

hs hs OF CANDIDATE PREDICTORS AND CONFOUNDERS AND PREDICTORS CANDIDATE OF

log 435) (Screening

ng remaining eligibility remaining ng

: The the through The : the of prediction flow participants (n=

SSEMENT

ASSE

Other reasons (n=14) Other reasons Declined to participate (n=8) participate to Declined

meeti Not

EXCLUDED EXCLUDED

recruitment and training of new trial new of training and recruitment

newly developed arm arm newly developed

diagnosed with IS or HS or IS with diagnosed

All patients consecutively admitted to the stroke rehabilitation uni rehabilitation stroke the to admitted consecutively patients All

for the initial 5 mont 5 initial for the

Patients with a newly developed upp developed a newly with Patients

•

Figure 5 Figure stu

27 Intervention

Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) Both the intervention and the control group received 1 hour of daily ESS to the affected arm from Monday to Sunday throughout the hospital stay, but not longer than 4 weeks after stroke onset. The ESS-intervention was initiated as soon as possible within the first 7 days post-stroke. Two sets of external electrodes were placed at the wrist, elbow and shoulder. The electrode at the wrist targeted all three peripheral nerves; the stimulation at elbow and shoulder was of cutaneous art. The intervention group received suprasensory (i.e. highest current amplitude without discomfort, pain or muscle twitches) ESS delivered in continuous mode (pulse width=250 µs, frequency=10 Hz) (active ESS). The control group received suprasensory ESS delivered in intermittent mode (active stimulation intervals of 3s are delivered in loops of 2.5 min, pulse width=250 µs, frequency=10Hz) (placebo ESS). The difference between the active and placebo ESS consisted in the amount of the delivered active ESS; the control group received a total dose of only 2% of the amount of active ESS received by the intervention group per ESS-session. Ideally, we would have preferred to employ a completely neutral placebo (i.e. no electrical stimulation), but concerns regarding a possible high drop-out percent in the control group made us design a placebo with a very low dose of active ESS, and therefore, with presumably no treatment effect. Our concerns regarding a high drop-out percent in the control group were caused by the fact that suprasensory ESS-treatment is perceivable by nature. Because the ESS- trial was conducted as a single-centre study, the participants could easily interact to each other at the stroke rehabilitation unit. Therefore, there was a considerable risk that participants in the control group could easily figure out that they did not get any active ESS-treatment and consequently lose their motivation for staying in the trial. Figure 6 visualizes the difference between active and placebo ESS.

28 Figure 6: A single ESS-session: the difference between active and placebo ESS

The active/placebo ESS was immediately followed by a minimum of 15 minutes’ training of the affected arm within the first 30 minutes after ESS’s cessation. We expected that the brain excitability would be increased during this time interval due to ESS (56, 60). The arm training consisted of active, repetitive, task-oriented practice. If the trial participant presented a severe arm paresis, the treating therapist decided which intervention methods to employ. A task- oriented exercise bank (1) was available for the therapists; the arm training was not further standardized. The active/placebo ESS was provided supplementary to: a) usual rehabilitation (i.e. PT and OT training delivered in the hospital’s stroke unit, community rehabilitation centres, and nursing homes), and b) other physical stroke rehabilitation services (e.g. PT services purchased by the participants themselves in private clinics) during the 6-months participation in the trial.

Figure 7: The ESS-intervention

29 Assessment instruments

Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) Figure 8 provides an overview of the outcome measures used in the ESS-trial.

Figure 8: Overview of the outcome measures used in the ESS-trial

6 4

7 5 3

1: Box and Block Test (BBT); 2: Pain during performance of BBT (NRS); 3: The upper extremity section of the Fugl-Meyer Assessment, subscale A−D (FMA-UE-AD); 4: Hand grip strength; 5: Pinch strength (palmar, lateral, key) 6: Perceptual threshold of touch (PTT); 7: modified Rankin Scale (mRS)

The Box and Block Test (BBT) (84) at 6 months post-stroke was the primary endpoint in the ESS-trial. The BBT is one of the most prevalent upper extremity outcome measures in stroke trials (85) and quantifies unilateral gross hand dexterity, which is defined in this PhD-thesis as the ability of reaching, grasping, moving and releasing objects. These motor skills are critical components of ADL-performance and, consequently, they are of key relevance when evaluating the effect of rehabilitation interventions on the recovery of arm functioning. Using the ICF- terminology, the BBT is a measure of the arm’s capacity of performing activities in a standard environment, and not a measure of what the person is doing in real life situations. The test result is the number of blocks that the person carries over the box partition from one compartment to the other during a 1-minute trial. Thus, higher numbers of blocks/min express better hand dexterity. The BBT has demonstrated excellent inter- and intra-observer reproducibility (86, 87), 30 and it has been validated for use in stroke patients (87, 88). It should be noted that a floor effect has been reported in patients with neurological disorders (87), and the minimally clinically important difference (MCID) has not been established yet. The minimal detectable change (MDC) is estimated to 5.5 blocks/min (86). Normative values are available for the healthy adult population (84). The BBT takes 2−5 minutes to administer, standardized test instructions are available, and it does not require other assessor training than reading the instructions (84, 89). In the ESS-trial, we administered the test according to the published protocol (84); one trial was performed for each arm, starting with the less affected one.

The secondary outcome measures were: a) pain in the arm during performance of BBT using Numerical Rating Scale-11 (90). b) the upper extremity section of the Fugl-Meyer Assessment (FMA-UE). According to a recent systematic review (85), the FMA-UE (91) is the most frequently used upper extremity outcome measure in stroke intervention studies. The FMA have a lower and an upper extremity section; these sections are often used separately. The FMA-UE includes various items, primarily measuring the ICF-domain of body functions. Based on the clinician’s observation of the patient’s performance, each item is scored on a 3-level ordinal scale (0/1/2); higher scores denote lesser impairment. The items are grouped in following subscales, each subscale presenting a sum score of the comprising item scores: • Subscale A covers reflexes, and ability of performing volitional gross arm movements (0−36 points); • Subscale B covers the ability of volitional movements of the wrist (0−10 points); • Subscale C covers the ability of volitional finger movements and grasping objects (0−14 points); • Subscale D covers coordination (0−6 points); • Subscale H covers sensory function (0−12 points); • Subscale J covers passive joint movements and joint pain (48 points). It has been shown that stroke patients with an impairment level < 31 points on FMA-UE subscales A−D (FMA-UE-AD, 0−66 points) are unlikely to be able to perform simple daily tasks, such as drinking from a glass, with their affected arm to (92, 93). A minimum of 58 points has been suggested to represent the lower limit for an almost complete arm recovery (94). Among the strengths of the FMA, its convincing and extensively investigated psychometric

31 properties are noteworthy (95, 96). FMA has an excellent inter- and intra-observer reproducibility and validity. In chronic stroke patients with mild-moderate motor impairments, improvements of 4−7 points on FMA-UE-AD are likely to be clinically meaningful (97). In patients with subacute stroke and severe motor impairments, the estimated MCID is 9−10 points on FMA-UE-AD (98). Regarding limitations, the FMA-UE presents a ceiling effect (96) and is considered too complex and time-consuming to administer (30-40 minutes) (99). The clinician is required to read the instructions on the test sheet or in a recently proposed manual (100). In the ESS-trial, we conducted the assessment according to the instructions on the English version of the test sheet (101). For each item, the less affected arm was tested first. If the assessor was in doubt about the quality of the observed performance, the patient was asked to perform the items 1−2 additional times (i.e. a total of 1−3 trials per item for each arm); the best item-performance was recorded. c) Maximal hand grip strength and 3 types of pinch strength (palmar, key, tip). Maximal hang grip strength, and maximal palmar and key pinch have shown an excellent test- retest reproducibility in patients with stroke (86). MCID is estimated to 5kg for the strength of the affected dominant hand, and 6.2kg for the affected non-dominant hand (102). MDC is 1.2kg for palmar pinch and 1.4kg for key pinch. Hand grip and pinch strength were measured using the digital hand dynamometer (model DHD-1) and the digital pinch gauge, respectively, from Saehan Corporation (www.saehanmedical.com). In the ESS-trial, the testing protocol was slightly adjusted to meet the requirements of the stroke population. The participants were sitting, had the shoulder adducted and the elbow flexed at 90 degrees and the forearm and wrist in neutral position. During the hand grip test, the participant’s forearm was not lying on the armrest, but the assessor was supporting the patient’s hand and the lower part of the dynamometer with one hand, and the upper part of the device with the other hand. During the pinch test, the patient’s forearm was lying on a table in front of the patient, and the shoulder was slightly flexed and internal rotated; the assessor supported the pinch gauge with one hand without grasping it; if necessary, the patient was asked to hold the forearm in neutral position using the less affected hand. Lateral and key pinch were tested with the thumb on top, and key pinch with the index finger on top. Participants were asked to squeeze as hard as possible and verbal encouragements were given each time. Between 3 to 5 measurements were performed per assessment; though, if arm paralysis, only one measurement was undertaken. Only the maximal value of the recorded measurements per each assessment was entered in the analyses. The

32 assessor supported the upper part of the dynamometer with one hand, and the lower part of the dynamometer with the opposite hand. d) perceptual threshold of touch (PTT). PTT refers to the minimal stimuli level of touch that is discernible in humans (103) and is a measure of sensory function. Eek et al (104) have demonstrated that PTT in the hand can be assessed with an excellent intra- and inter-rater reliability (ICC: 0.96−0.99) in elderly stroke patients using an easy-to-follow transcutaneous electric nerve stimulation (TENS) protocol. Recently, normative values of PTT in the hands in relation to age, and sex and right/left side have been established for the healthy adult population (105). The advantage of using electrical stimulation in assessing the PTT rather than Semmes-Weinstein monofilaments is the generated numerical values (assessed in mA) that allows monitoring changes over time accurately. In the ESS-trial, the assessment of PTT was performed according to the protocol using the CEFAR Tempo device and self-adhesive TENS/NMES/FES electrodes (3.2cm) from DJO Nordic AB (Murmansgatan 126, 212 25-Malmö, Sweden). Between 3 to 5 measurements were conducted per assessment. Subsequently, the lowest and the highest values were discarded, and the value entered in the analyses was the mean of the remaining measurements. e) modified Rankin Scale (mRS). The mRS (106) is of the most frequently used functional outcome measures in stroke trials (107). Based on the patient’s independence in mobility and in performance of ADLs, the clinician assesses a global score of disability using an ordinal scale with levels ranging from 0 to 6. A value of mRS=0 means no left symptoms at all; mRS≤2 denotes being able to live independently despite some remaining symptoms or inability to resume some complex ADLs; mRS≤4 means not requiring constant assistance from a caregiver; mRS=6 is used if the patient is dead. Various methodologies have been employed for collecting the background information, e.g. direct interview, telephone interview, structured interview or through medical records (106, 107). The mRS has been criticized for poorly defined boundaries between the disability levels, resulting in a high inter-rater variability (108). Efforts have been made to minimize this limitation of the instrument by developing a structured interview (109, 110), a simplified version of the interview/questionnaire (smRSq) (111), and a web-based training and certification course (112). The intra-rater reliability has been proven to be excellent. Other limitations include poor responsiveness to change, and a floor effect at admission to rehabilitation (113). A MCID≥1 point has been proposed (113). In the ESS-trial, the information used to assess the mRS was 33 collected directly from the patients, and to a lesser extend from their nursing staff, relatives and medical records; the scale from 0 to 5 was employed and the values for dead patients was left blank in the data set. The assessor (EG) was web-certified user of mRS.

Data on all outcome measures were collected at: a) baseline (within the first week after stroke onset); b) at post-intervention (at hospital discharge or latest 4 weeks post-stroke); and c) at 6- months follow-up (6 months after stroke onset). All baseline measurements were performed at the stroke unit. Post-intervention and follow-up assessments were performed at the participants’ residences at that specific point in time (stroke unit, patient’s home, inpatient rehabilitation centers, nursing homes). All assessments were performed by the same assessor (EG). All outcome measures were regarded as numerical scales. Although we collected data on both the affected and the less affected arm, only data on the affected arm are presented in this PhD-thesis.

Paper III: Early predictors of the affected arm functioning at 6 months post-stroke The predicted outcome was two different levels of FMA-UE-AD at 6 months post-stroke: a) FMA-UE-≥ 32 points, corresponding to a minimum function level that allows the patient to perform basic ADL with the affected arm (92, 93), and b) FMA-UE-≥ 58 points, corresponding to mild level of impairment (94) and a high probability that the patients routinely use the affected arm in performance of ADL (22).

We examined the individual predictive value of following items form FMA-UE-AD assessed during the first week post-stroke: a) shoulder abduction within synergies (subscale A.II), b) elbow extension within (subscale A.II), c) forearm pronation/supination (subscale A.III), d) wrist stability of 15 degrees dorsiflexion (subscale B), e) mass finger extension (subscale C), f) pincer grasp (subscale C), and g) cylinder grasp (subscale C); all ordinal variables with possible values of 0 (no function), 1 (partial function) or 2 (full function). Likewise, we considered the prediction value of the sensory function measured with FMA-UE-H and dichotomized as full sensory function (score =12) and sensory dysfunction (score <12).

34 Sample size

Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) A pre-trial sample calculation showed that a total of 102 participants were required to detect a within-group improvement of 5.5 blocks/min (=MDC) (86) on BBT with a 2-sided significance level of 5% and a power of 80%; a possible drop-out of 28% was included.

Paper III: Early predictors of the affected arm functioning at 6 months post-stroke The size of the cohort was generated by merging the SALGOT (n=122) and the ESS (n=102) samples. For further details, see Figure 5.

Randomization

Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) The study participants were allocated sequentially either to the control or the intervention group using a randomization list constructed by block randomization with variable block size. The randomization list was stratified by sex and ability to extend finger voluntarily at baseline (25), and generated with the random generator in SAS version 9.4. The assignment of participants to trial groups was primarily conducted by administrative staff that was not involved in the trial. To a lesser extent, the senior investigator (SPM) was involved when the administrative staff was absent.

Blinding

Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) The study participants were kept unaware of which stimulation mode was the intervention and which one was the control. Because suprasensory ESS is perceivable and because we could not prevent the participants from interacting with each other during their hospital stay, a complete blinding of the participants was not achievable. The therapists responsible for usual rehabilitation, the medical staff, the outcome assessor (EG) and the data analysts (VS, EG) were blinded to the group allocation until analyses were completed.

35 Statistical methods

Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) Fischer’s Exact Test (for categorical variables) and Wilcoxon Non-Parametric Test (for continuous variables) were used to compare the trial groups with respect to demographic and clinical characteristics at baseline, the achieved ESS-intervention, adverse reactions to ESS, complications during hospital stay, length of hospital stay, discharge destination and residence at 6-month follow-up.

The development of the outcome variables was analyzed in longitudinal models over the first 6 months after stroke onset (i.e. baseline, post-intervention, follow-up). At each time point, the differences in outcomes between the trial groups were calculated using multivariable regression analyses. Analyses were adjusted for the stratification variables (i.e. sex and ability to perform voluntary finger extension at baseline). The available data at each time point was weighed with the inverse probability of being observed to adjust for possible differential dropout; these probabilities were estimated from logistic regression models with sex, affected dominant hand, pre-stroke mRS and outcomes at previous trial time points as covariates. A visual inspection of the summary statistics of the missing participants showed that they were primarily women, had the dominant hand affected and a pre-stroke mRS>0. Because mRS data at baseline were not collected, we assigned a baseline mRS of (4+5)/2=4.5 points to all the trial participants. This estimation is supported by the assumption that hospitalized stroke patients have a mRS of 4 or 5 points due to their need for medical attention.

Paper III: Early predictors of the affected arm functioning at 6 months post-stroke We employed Wilcoxon Non-Parametric Test and Fisher’s Exact Test to determine whether the outcome groups (FMA-UE-AD< 32/ ≥32, FMA-UE-AD< 58/≥58) were statistically different with respect to demographics and clinical characteristics at baseline, and other confounders. For each potential predictor we used logistic regression to calculate the odds ratio for a favorable outcome (FMA-UE-AD≥ 32 and ≥ 58) among patients that had partial or full volitional motor function (item score: 1 or 2), or full sensory function (FMA-UE-H=12) in comparison with the group of patients that presented no volitional motor function (item score=0) or sensory dysfunction (FMA-UE-H<12); the regressions were weighed as described below. Three logistic regression analyses were performed for each predictor: a) unadjusted; b) adjusted for following confounders: sample (SALGOT/ESS), sex, age and living arrangement, and c) adjusted for following confounders: sample, age, sex, living arrangement, previous stroke, stroke diagnosis, 36 affected dominant hand, leg paresis, aphasia, stroke severity (Scandinavian Stroke Scale, SSS) (114, 115), pre-stroke physical activity level measured with Saltin-Grimby Physical Activity Level Scale (SGPALS-4) (116, 117), number of hospital days, and number of days from the stroke onset to the measurement of the potential predictors (see Table 6). The SALGOT-study assessed the stroke severity using the National Institutes of Health Stroke Scale (NIHSS-scale) (118), and the pre-stroke physical activity level with SGPALS-6 (119, 120). For the purpose of this study, the NIHSS-values were converted into SSS-values using the mathematical equation SSS=50.37–1.63xNIHSS (121); categories 1 and 2 of SGPALS-6 were merged into category 1 of GSPALS-4, and categories 5 and 6 of SGPALS-6 were merged into category 4 of SGPALS-4.

Potential bias because of differential dropout and missing values was countered by weighting the remaining observations with the inverse of the estimated probability of this value being observed. These estimated probabilities were for each observation estimated from a logistic regression model on the observation being missing or not including all the above-mentioned confounders.

In both studies, the statistical significance level was set to 1% because of multiple comparisons; in the ESS-trial, though, a significance level of 5% was applied in comparisons regarding the primary outcome. Analysis was performed with SAS version 9.4.

Key results

Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial) Baseline characteristics of trial participants (n=102) were collected from medical records. As shown in Table 1, the intervention and the control group were well balanced at baseline due to the randomization process. The median age was 71 years in the control group and 72 years in the intervention group (71/72 years); men and women were equally distributed in the groups; 43/45% of the participants were current smokers; 47/51% had increased-to-high risk of metabolic disease; 53/62% had hypertension; both groups had a lower median pre-stroke physical activity level (PASE=76/88 points) than the proposed median value for healthy individuals aged 65-100 years (PASE=90 points) (122); 50% of the participants had some insignificant-to-slight disability before current stroke (pre-stroke mRS Q1–Q3 = 0–2); 76/81% had an IS, mainly caused by small-artery occlusion; 92/100% of the trial participants had a mild-

37 to-moderate stroke (SSS>25 points); 76/79% were able to perform some voluntary finger extension at baseline; 96/100% had the right hand as dominant, and 53/47% had the dominant hand affected. Participants presented a moderate arm impairment level (FMA-UE-AD=37, median) (94) (see Table 2).

38 Table 1: Baseline demographic and clinical characteristics of the participants in the ESS-trial (n=102)

Control Intervention p- group (n=49) group (n=53) value Demographic characteristics Age, years, median (Q1–Q3) 71(64–80) 72 (64–79) 0.93 Sex, men, n (%) Men 25 (51) 28 (52.8) 1.00 Women 24 (49) 25 (47.2) Living arrangement, n (%) Living alone 28 (57.1) 34 (64.2) 0.54 Living with othersa 21(42.9) 19 (35.9) Actual profession/last profession before retirement, n (%) Non-manual workersb 29 (59.2) 33 (62.3) 0.65 Self-employedc 3 (6.1) 1 (1.9) Manual workersd 17 (34.7) 19 (35.9) Pre-stroke disability level (pre-stroke mRSe), mean 0.8 (1.1) 1 (1.3) 0.41 (SD), median (Q1–Q3) 0 (0–2) 0 (0–2) Stroke risk factors Smoking, n (%) Never smoker 8 (16.3) 15 (28.3) Former smoker 20 (40.8) 14 (26.4) 0.22 Current smoker 21 (42.9) 24 (45.3) High risk of disease due to alcohol consumptionf, 10 (20.4) 9 (17) 0.80 n (%) Increased to high risk of metabolic diseaseg, 22 (46.8) 25 (51) 0.69 n (%) Previous stroke, n (%) 12 (24.5) 10 (18.9) 0.63 Previous transient ischaemic attack, n (%) 1 (2) 1 (1.9) 1.00 Previous atrial fibrillation, n (%) 10 (20.4) 8 (15.1) 0.61 Previous myocardial infarction, n (%) 1 (2) 1 (1.9) 1.00 Previous angina pectoris, n (%) 0 (0) 0 (0) - Diabetes, n (%) 8 (16.3) 7 (13.2) 0.78 Psychiatric disorder, n (%) 3 (6.1) 4 (7.6) 1.00 Stroke in the family, n (%) 2 (4.1) 5 (9.4) 0.44 Heart failure, n (%) 8 (16.3) 3 (5.7) 0.11 Hypertension, n (%) 26 (53.1) 33 (62.3) 0.42 Myocardial infarction in the family, n (%) 2 (4.1) 3 (5.7) 1.00 Peripheral arterial disease, n (%) 1 (2) 1 (1.9) 1.00 Hyperlipidaemia, n (%) 9 (18.4) 15 (28.3) 0.25 Other diseases, n (%) 18 (36.7) 16 (30.2) 0.53 Pre-stroke physical activity level (PASEh), 76 (40–134) 88 (40–149) 0.58 median (Q1–Q3) Clinical characteristics Diagnosis, n (%) Ischaemic stroke (IS) 37 (75.5) 43 (81.13) 0.63 Haemorrhagic stroke (HS) 12 (24.5) 10 (18.87) TOASTi classification of subtypes of ischemic stroke, n (%) Large-artery artherosclerosis 4 (10.8) 6 (14) Cardioembolism 9 (24.3) 8 (18.6) Small-artery occlusion 24 (64.9) 28 (65.1) 0.86 Stroke of undetermined etiology 0 (0) 1 (2.3) Stroke of other determined etiology 0 (0) 0 (0) Stroke severity (SSSj), n (%) Mild to moderate stroke (SSS>25) 45 (91.8) 53 (100) 0.05 39 Major stroke (SSS≤25) 4 (8.2) 0 (0) Voluntary finger extension, yes, n (%) 37 (75.5) 42 (79.3) 0.81 Affected upper limb, right, n (%) 28 (57.1) 25 (47.2) 0.33 Dominant hand, right, n (%) 47 (95.9) 53 (100) 0.23 Affected dominant hand, n (%) 26(53.1) 25(47.2) 0.69 Sensory dysfunctions, (FMA-UE-Hk), yes, n (%) 26 (57.8) 28 (52.8) 0.69 Joint pain, (FMA-UE-Jl), yes, n (%) 17 (37) 20 (37.7) 1.00 Acute treatment, n (%) Thrombolysis 6 (12.3) 9 (17) Aspirin 26 (53.1) 34 (64.2) Thrombectomy 0 (0) 0 (0) 0.08 Other 5 (10.2) 0 (0) Not relevant (for HS) 12 (24.5) 10 (18.9) Medications at hospital admission, n (%) Aspirin 8 (16.33) 9 (17) 1.00 Persantin 2 (4.08) 0 (0) 0.23 Anticoagulation medications 8 (16.33) 6 (11.3) 0.57 Clopidogrel 6 (12.24) 9 (17) 0.58 High blood pressure medications 29 (59.18) 32 (60.4) 1.00 Lipid lowering medications 12 (24.49) 14 (26.4) 1.00 Other 36 (73.47) 34 (64.2) 0.39 Blood pressure (BP) (mmHg), median (Q1–Q3) Systolic BP (SBP) 148 (135–170) 153 (135–171.5) 0.56 Diastolic BP (DBP) 78 (69–87) 81.5 (72–89) 0.35 Pulse (/min), median (Q1–Q3) 71 (62–80) 76.5 (62–82.5) 0.41 Blood tests Total cholesterol ≥ 5.0 mmol/l 20 (46.5) 33 (67.4) 0.06 eGfR ≤ 60 ml/min 14 (30.4) 15 (28.3) 0.83 Glucose ≥ 7.7 mmol/l 10 (23.8) 6 (13.3) 0.27 aMarried, house share, sheltered housing, nursing home. bBusiness executives, engineers with university degrees, physicians, college teachers, secondary school teachers, office assistants, sales people (123) cProfessionals with and without employees, entrepreneurs, farmers (123) dAuto mechanics, metal workers, construction workers, factory workers, waiters, cleaning staff (123) emodified Rankin Scale 0–5 (106) f(Sex=male AND alcohol consumption > 21 drinks/week) OR (Sex=female and alcohol consumption > 14 drinks/week) (124) g(Sex=female AND 25≤BMI<30 AND waist circumference<88 cm) OR (Sex=male AND 25≤BMI<30 AND waist circumference<102 cm) OR (Sex=female AND 25≤BMI<30 AND waist circumference≥88 cm) OR (Sex=male AND 25≤BMI<30 AND waist circumference≥102 cm) OR (Sex=female AND BMI≥30) OR (Sex=male AND BMI≥30). Waist circumference measured in standing position (125) hPhysical Activity Scale for Elderly (122) iClassification of subtype of acute ischemic stroke (126) jScandinavian Stroke Scale (114, 115) kUpper extremity section of Fugl-Meyer Assessment, subscale H (Sensation, 0-12) lUpper extremity section of Fugl-Meyer Assessment, subscale J (Joint pain during passive movement, 0−24) *p-value≤0.01

40 All randomized participants (n=102) were included in the intention-to-treat analysis. We demonstrated no statistically significant between-group differences for any outcome measure and at any time point (i.e. baseline, post-intervention, follow-up) (Table 2). Figure 9 visualizes the same main results in graphical form as the development of all outcome measures (means weighted for possible differential dropout) throughout the course of the trial. Baseline assessments were conducted within 7 days post-stroke; post-intervention assessments at a median of 19−20 days after stroke and follow-up assessments at a median of 3.5−5 days after the actual 6-month date post-stroke.

]

], ],

4.1

;

; 0.9],

2.5; 1.9],

6.7; 5

0.6; 0.3],

0.7; 0.4],

1.1; 0.5

4.3; 1.2],

5.2

)

0.3;0.8],

[

ollow

value

0.2

0.1 [

0.3

1.6 [

0.3

0.9 [

0.5

p=0.49

p=0.62

p=0.38

0.2 [

p=0.88

p=0.44

p=0.27

p=0.80

p=0.77

p=0.82

extremity Fuglof section

2.3

Effect size (

95% CI

[

0.4],

ost

Adjusted difference

; 0.4], ;

2.2; 2

pper

;

0.9; 0.2],

2.1; 1.1],

3.9;

0.1; 0.7],

0.5; 0.5

3.2;6.7],

[

0.3

erceptual thresholderceptual touch;of

[

0.3 [

0.3[

0.1 [

0.5

0.8

AD: U AD:

p=0.82

p=0.20

p=0.12

0.3 [

p=0.44

p=0.93

0.0 [

p=0.92

p=0.53

p=0.49

1.7 [

p=0.62

)

24.4)

3.4)

6.9)

4.4)

63)

–

FMA

–

(1.8

(2.9)

2.5

0.3 (1

2 (2

0 (0

2.3 (1.6)

3.1 (2.3)

18.1(10.5)

58 (46

27 (16

25.9(15.7

50.7(17.2)

3.3 (2.6

2.2 (1.0

4.5 (2.2

3.1 (1.6

intervention; PTT: P

17.7(9.9

(I)

Q3)

)

–

)

59)

group

5.6

–

Post: Post

2.7)

4.0)

28)

22.1)

–

(2.3

(36

(19.5)

11;

Post

ean (SD) ean

3 (3

0 (0

3.1 (0.9

0.3 (1.5)

1.8 (1.6)

3.7 (2.6)

2.3 (2.1)

4.2

1.9 (0

2.4 (0

14.9(10.9)

3.8 (2.8

19.2(15.2

edian (Q1

49.5

3.7 (1.4

15.1(6

; BBT: and Box Block Test;

Intervention

)

4.5

)

50)

–

16.1)

2.5)

4.3)

2.4)

–

(4.6

(15

4.5 (0

2.3 (3

0 (0

9.3 (11

0.3 (0.9)

1.3 (1.4)

1.4 (2.4)

5.6

3 (0

Baseline

11.0(10

0.6 (0

1.3 (0

0.7 (0

4.5 (3.5

33.6(19.1)

4.5 (4.5

8.8 (2.6

5); Numerical NRS: Scale

−

)

64)

e (0

)

28.7)

4.5)

7.1)

5.4)

4.3)

–

(3.2

(49.5

2.4

5.0 (3

2 (2

0.5 (1.9)

2.7 (1.9)

3.3 (2.5)

4.6

18.4(13.4)

28 (14

26.7(16.7

51.2(18.4)

2.7 (1.1

4.5 (2.5

3.5 (1.5

3.5 (2.9

16.6(6.5

(C)

)

Q3)

–

)

odified Rankin Scal

)

25.4

4.1)

5.9)

5.0)

62)

4.9)

32)

–

group

–

(0.9

(5.4

ean (SD) ean

(38

4.0 (3

3 (2

0 (0

3.2

0.2 (0.9)

2.4 (2.1)

2.9 (2.7)

5.1

23 (3

edian (Q

2.3 (0

16.5(12.8)

19.9(16.2

46.7(20.6)

mRS: M mRS:

2.1 (0.5

4.0 (0.9

3.3 (2.8

14.8(5.1

Control

comparisons

)

4.5

7.4)

55)

2.6)

3.2)

–

17)

4.3)

–

djusted and for abilitysex of voluntary finger extension at baseline

(6.7

gro

4.5

0 (0

2.7 (3

0.6 (1.7)

1.4 (1.8)

1.9 (2.4)

1 (0

Baseline

2 (0

12.0(13

33.2(21

0.5 (0

37 (10

10.2(12.8

7.1 (0.7

4.5 (4.5

4.5 (3.5

ment, subscale A−D;

)

Between

(

r Assess

points

pinch pinch

)

−

trength

blocks/min

*p≤0.01

Meye

Adjusted difference: A

Table 2 Table

mRS

(0−10 points)

NRS

Tip pinchTip

strength

Key

strength

Palmar Palmar pinch

Hand grip

(

PTT

(0−66 points)

FMA ( BBT

42 Figure 9: The development of all outcome measures (weighted* means with 95% CI) throughout the course of the trial

(blocks/min)

)

(0−66 points (0−66

)

(kg

)

(mA

) 10 -

0 (points

)

- (points 0 (points

*Weighted for sex, dominant hand affected, pre-stroke mRS and outcomes at previous trial points; BBT: Box and Blocks Test; FMA AD, FMA-UE-AD: Upper extremity section of Fugl-Meyer Assessment, subscale A−D; NRS: Numerical Scale-11; Pre: Baseline (for mRS, the baseline values are the pre-stroke values; the baseline values estimated at 4.5 points are not depicted); Post: Post-intervention; PTT: Perceptual threshold of touch.

Despite the lack of statistically significant differences between the intervention and the control group, there were statistically significant improvements in both trial group on all outcome measures, except for pain, throughout the whole trial (Table 3); improvements in FMA-UE-AD and mRS were moreover ≥MCID at all time points, and for the hand grip strength at follow-up (see Table 3 and Figures 10−15).

46 Table 3: Within-group differences

Control group Intervention group Adjusted difference Adjusted difference [95% CI], p-value [95% CI], p-value Baseline to Baseline to follow-up Baseline to Baseline to follow-up post-intervention post-intervention BBT 8.9 [ 6.8; 11] 16.0 [12.7; 19.1] 9.6 [ 7.4; 11.9] 16.5 [13.1; 19.9] (blocks/min) p<.0001* p<.0001* p<.0001* p<.0001* FMA-UE-AD 12.4 [ 8.3; 16.5] 16.6 [12.6; 20.6] 10.7 [8; 13.4], 17.5 [ 13.2; 21.8] (0−66 points) p<.0001* p<.0001* p<.0001* p<.0001* PTT -2 [-3.4; -0.6] -2.3 [-4.1; -0.6] -1.4 [-2.4; -0.6] -2.0 [-3.4; -0.7] (mA) p=0.003* p=0.008* p=0.001* p=0.002* Hand grip 3.7 [2.2; 5.3] 5.6 [3.9; 7.4] 3.8 [2.5; 5.2] 7.2 [5; 9.4] strength (kg) p<.0001* p<.0001* p<.0001* p<.0001* Palmar pinch 0.8 [0.5; 1.2] 1.3 [0.8; 1.8] 0.9 [0.5; 1.2] 1.6 [1; 2.2] strength (kg) p<.0001* p<.0001* p<.0001* p<.0001* Key pinch 1 [0.5; 1.5] 2.1 [1.5; 2.6] 1.3 [0.7; 1.8] 2.1 [1.4; 2.9] strength (kg) p<.0001* p<.0001* p<.0001* p<.0001* Tip pinch 0.8 [0.5; 1.1] 1.1 [0.7; 1.5] 0.5 [0.2; 0.7] 1 [0.4; 1.3] strength (kg) p<.0001* p<.0001* p<.0001* p<.0001* NRS -0.3 [-0.6; 0] -0.1 [-0.5; 0.3] 0 [ -0.4; 0.5] 0.1 [-0.3; 0.4] (0−10 points) p=0.076 p=0.723 p=0.850 p=0.728

mRS -1.3 [-1.6; -1.1] -2.1 [-2.4; -1.8] -1.4 [-1.6; -1.1] -1.9 [-2.2; -1.6] (0−5 points) p<.0001* p<.0001* p<.0001* p<.0001* Adjusted difference: Adjusted for sex and ability of voluntary finger extension at baseline; BBT: Box and Blocks Test; FMA-UE-AD: Upper extremity section of Fugl-Meyer Assessment, subscale A−D; NRS: Numerical Scale-11; PTT: Perceptual threshold of touch; mRS: Modified Rankin Scale; *p≤0.01; Differences ≥ MCID (if established) are in bold.

Figure 10: Within-group adjusted* differences for BBT (*adjusted for sex and ability of voluntary finger extension at baseline)

Control gr. Figure 11: Within-group adjusted* differences for FMA-UE-AD (*adjusted for sex and ability of voluntary finger extension at baseline)

Figure 12: Within-group adjusted* differences for hand grip strength (*adjusted for sex and ability of voluntary finger extension at baseline)

48 Figure 13: Within-group adjusted* differences for palmar pinch strength (*adjusted for sex and ability of voluntary finger extension at baseline)

Figure 14: Within-group adjusted* differences for key pinch strength (*adjusted for sex and ability of voluntary finger extension at baseline)

49 Figure 15: Within-group adjusted* differences for mRS (*adjusted for sex and ability of voluntary finger extension at baseline)

Despite improvements ≥MDC=5.5 blocks/min in BBT from baseline to 6 months post-stroke, the dexterity level of the participants in the ESS-trial was far below the reported normative values (84). The female participants in the ESS-trial (n=49) had a median age of 74 years and presented a hand dexterity of 26 blocks/min on BBT at 6 months post-stroke; the male participants (n=53) were in average 68 years of age (median) and their final dexterity level was of 27 blocks/min. Both women and men exited the ESS-trial with a dexterity level of approximately 40% of the reported dexterity level for healthy, age- and sex-matched individuals (68-69 blocks/min for women, 67-68 blocks/min for men, depending on whether the affected hand is right or left) (84). Thus, they were still demonstrating a considerably decreased capacity for arm activity at 6 months after stroke, when the recovery of arm functioning is expected to plateau. Notably, the global disability level of the trial participants at 6 months after stroke was approximately 2–3 points (median: 2.4/2.5) on the mRS, suggesting that about 50% of the participants were independently mobile and likely to manage basic ADL without assistance (Table 2). This was in line with the fact that 79/80% of the participants were living at home 6 months after their stroke onset (Table 4).

Table 4 presents additional results, and Table 5 gives a detailed description of the achieved ESS- intervention. No statistically significant differences were found between the trial groups.

50 Table 4: Additional results

Control group Intervention group p-value (n=49) (n=53) Adverse reactions to ESS, n (%) 1 (2) 1 (1.9) 1.00 Length of hospital stay, days, median (Q1―Q3) 17 (12―23) 18 (15―22) 0.69 Recurrent stroke during participation in the study, n (%) 3 (6.1) 2 (3.8) 0.67 Complications during the hospital stay, n (%) Pneumonia 8 (16.3) 2 (3.8) 0.05 Urinary tract inflammation 14 (28.6) 14 (26.4) 0.83 Deep vein thrombosis 1 (2) 2 (3.8) 1.00 Pulmonary embolism 0 (0) 1 (1.9) 1.00 Fall 2 (4.1) 1 (1.9) 0.61 Discharge destination, n (%) Home 19 (38.8) 24 (45.3) Outpatient rehabilitation center in the community 1 (2) 1 (1.9) Inpatient rehabilitation center in the community 25 (51) 28 (52.8) 0.52 Nursing home, sheltered housing 1 (2) 0 (0) Hospital 1 (2) 0 (0) Dead 2 (4.1) 0 (0) Residence at follow-up, n (%)* Home 37 (78.7) 41 (80.4) Outpatient rehabilitation center in the community 0 (0) 0 (0) Inpatient rehabilitation center in the community 2 (4.3) 0 (0) 0.53 Nursing home, sheltered housing 4 (8.5) 6 (11.8) Hospital 0 (0) 1 (2) Dead 4 (8.5) 3 (5.9) p≤0.01

value

0.32

0.6

0.52

1.00

0.82

0.75

0.29

0.33

0.6

0.0

0.33

0.84

0.2

0.1

0.80

0.45

1(0―2)

1 (1―3)

2 (1―4)

12 (23.5)

5 (1―16)

13 (9―18)

10 (3―13)

11 (7―15)

9 (2―18.8)

20 (15―26)

13 (10―20)

group (n=53)

Intervention

7.1 (5.3―9.5)

5.3 (4.3―6.1)

27 (1.80―39.5)

15.7(6.6―24.4)

34.4(16.4―47.7)

1(0―1)

Control

1 (1―4)

2 (1―4)

10 (23.3)

9 (5―13)

4 (3―5.6)

12 (8―16)

11 (6―15)

13 (8―18)

3.5 (0―14)

19 (13―26)

group (n=49)

5.8 (4.6―9.5)

6.2 (0.9―19.8)

10.1(4.9―25.1)

30.9(16.0―65.0)

24.9(10.5―56.3)

up up assessment

arm arm training

, median ,(Q1―Q3) median

minutes

(%)

up assessmentsup

intervention, usual rehabilitation (including arm timepost actual training),points for and (including intervention, rehabilitation usual

assessment

minimum 15 minimum

intervention assessment, (Q1―Q3)median

median (Q1―Q3)median

intervention

post

stroke stroke to the scheduled date for follow

intervention period, (Q1―Q3) median hours,

intervention period, hours, (Q1―Q3)median

intervention and follow

up assessments

session to

intervention period, hours, (Q1―Q3)median

arm trainingarm following ESS)

session followed by

sessions, days, (Q1―Q3)median

the ESS

during trial, the hours, (Q1―Q3)median

interventions*, days,

essions, days, (Q1―Q3) median

sessions, days, (Q1―Q3)median

escription of the achieved ESS achieved escription of the

interventions (per protocol + per not protocol), (Q1―Q3)days, median

: D

intervention = ESS

intervention

interventions givennot per protocol, (Q1―Q3)days, median

protocol ESS

ESS

p≤0.01

median (Q1―Q3)median

No. 6 days from of postmonths

No. last days from of ESS

No. days from of stroke onset to post

Actual time Actualpoints time for post

Additional physicalAdditional stroke rehabilitation during the trial, n

Total during OT/PT thetraining trial, hours, (Q1―Q3)median

OT/PT after training

OT/PT during thetraining ESS

Usual rehabilitation (including training) arm

Total arm trainingTotal arm

Arm trainingArm after ES the

Arm trainingArm during the ESS

Arm trainingArm (including

Total ESS

ESS

Per

Partially completedPartially ESS

Completed ESS

Potential ESS

ESS

intervention and follow and intervention Table 5 Table

52 Paper III: Early predictors of the affected arm functioning at 6 months post-stroke The potential predictors and confounders were assessed at day 3 (median), (Q1−Q3) = (3−5), range: 0−7 days post-stroke. The ESS-trial conducted the baseline measurements at day 5 (median), (Q1−Q3) = (4−6), range: 0−7 days after stroke. Participants in the SALGOT-study (82) were all assessed at day 3 post-stroke at baseline. The outcome measure (FMA-UE-AD) was collected for 176/223 study participants at 6 months post-stroke.

The demographic and clinical characteristics of the ESS-SALGOT-cohort at baseline, and other confounders are listed in Table 6. The median age was 71 years, and the proportion of men was 54%. Half of the participants were living alone. The most prevalent risk factors for stroke were: a) hypertension: 54%, b) overweight (BMI≥25): 42%, and c) physical inactivity before stroke onset: 25%. The most participants had a IS (82%) due to small-artery occlusion (43%), and presented a mild-to moderate stroke (87%). The arm function was moderately impaired (FMA- UE-AD=36, median); 63% had leg paresis and 22% aphasia. The median number of hospital days was 21. Furthermore, Table 6 presents the distribution of baseline demographic and clinical characteristics, and other confounders separately in two outcome groups (FMA-UE-AD< 32/ ≥32, FMA-UE-AD< 58/≥58) at 6 months post-stroke. The groups with a less favourable outcome (FMA-UE-AD<32/58) comprised a significantly higher percentage of participants with major strokes, leg paresis and more impaired arm motor function at baseline, and a longer hospital stay.

Table 7 and 8 show the main results of this study. Overall, the odds ratios (OR) for a favourable outcome (FMA-UE-AD≥ 32/58) were significantly higher among patients with partial/full motor function or full sensory function compared with patients with absent motor function or sensory dysfunction at baseline. Since the estimated effect sizes (OR a−c) are less accurate due to large 95% confidence intervals (95%CI), the lower limits of the 95%CI can be used as conservative estimates. Thus, when adjusting for all confounders (OR c), the probability of achieving a FMA- UE-AD ≥32 was at least 7.3-fold higher when the patients presented partial shoulder abduction within synergies compared with patients with no shoulder abduction. Whether forearm pronation/supination, wrist dorsiflexion, and grasping ability predict a FMA-UE-AD≥32 remains unanswered since the odds ratio could not be calculated due to the distribution of our data (i.e. there were no patients with a score of 2 on these motor items at baseline and a FMA-UE-AD< 32 at 6 months after stroke) (see Table 7). Moreover, the probability of achieving an almost complete arm recovery (FMA-UE-AD≥ 58) were at least 2.2−19-fold higher among patients showing partial elbow extension within synergies, forearm pronation/supination, wrist 53 dorsiflexion, and pincer and cylinder grasp without resistance, when OR were adjusted for all confounders (OR c). In patients with full function on these items, the probability of a FMA-UE- AD ≥58 at 6 months post-stroke was at least 5.7−36.8-fold higher compared with patients with no motor function at baseline. Patients with intact sensory function were at least 2.4-fold more likely to achieve a FMA-UE-AD≥32 and at least 2.3-fold more likely to achieve a FMA-UE- AD≥58 at 6 months after stroke compared with patients with sensory dysfunction at baseline.

value

1.000

0.714

0.263

0.256

1.000

0.140

0.239

0.585

0.064

0.780

1.000

0.792

0.323

0.694

0.684

1.000

0.428

0.538

0.437

0.213

78)

−

92)

≥ 58

−

(7)

12 (13)

6 (8)

34 (44)

18 (23)

13 (17)

77 (82)

17 (18)

16 (18)

45 (48)

26 (28)

21 (22)

1 (1)

44 (47)

7 (7)

4 (4)

8 (9)

1 (1)

4 (4)

15 (16)

1 (1)

44 (47)

50 (53)

40 (43)

54 (57)

(38

69 (62

stroke

80)

−

92)

< 58<

−

(12)

2 (3)

27 (49)

16 (29)

9 (16)

56 (75)

19 (25)

20 (29)

36 (48)

29 (39)

11 (15)

2 (3)

46 (61)

7 (9)

3 (4)

8 (11)

3 (4)

2 (3)

14 (19)

1 (1)

9 (12)

39 (52)

36 (48)

37 (49)

38 (51)

(34

72 (62

value

AD at 6 months AD post 6at months

1.000

0.565

0.821

0.461

0.354

0.842

0.233

0.525

0.457

1.000

0.648

1.000

0.208

0.613

0.805

1.000

0.753

0.852

0.191

0.972

FMA

≥32

17 (13)

6 (6)

50 (49)

23 (22)

18 (17)

103 (78)

29 (22)

31 (28)

66 (50)

44 (33)

28 (21)

2 (2)

68 (52)

11 (8)

5 (4)

13 (10)

2 (2)

4 (3)

22 (17)

2 (2)

12 (9)

64 (48)

68 (52)

64 (48)

68 (52)

(34−92)

69 (62−80)

< 32<

cohort at baseline, and other and cohort baseline, (n=223) at confounders

4 (11)

2 (7)

11 (36)

4 (14)

30 (81)

7 (19)

5 (14)

15 (41)

11 (30)

4 (11)

1 (3)

22 (59)

3 (8)

2 (5)

3 (8)

2 (5)

7 (19)

0 (0)

4 (11)

19 (51)

18 (49)

13 (35)

24 (65)

(55−90)

68 (62−78)

ESS

30 (13)

15 (8)

10 (6)

78 (43)

51 (28)

28 (15)

182 (82)

41 (18)

53 (25)

93 (42)

79 (35)

37 (17)

3 (1)

120 (54)

21 (9)

10 (4)

28 (13)

4 (2)

8 (4)

45 (20)

2 (1)

22 (10)

111 (50)

112 (50)

103 (46)

120 (54)

(26−95)

71 (63−80)

n (%) n

=1), (%) n

n (%) n

yes, yes,

n (%) n

(SGPALS

(min−max)

and clinical characteristics of the SALGOT of the characteristics and clinical

yes,

n (%)n

(%)

stroke

n (%) n

yes, yes,

n (%) n

yes, yes,

≥25), (%) n

yes, yes,

n (%) n

yes, yes,

artery occlusion

artery artherosclerosis

yes, yes,

e otherof determined etiology

iagnosis, (%) n

arrangement, (%) n

Thrombolysis

Stroke undeterminedof etiology

Strok

Small

Cardioembolism

Large

Ischaemic stroke (IS)

Haemorrhagic stroke (HS)

Living Living otherswith

Living aloneLiving

Women

Men

Acute treatment,Acute (%) n yes,

TOAST classification of subtypes of ischemic stroke,(%) n

Stroke d

Clinical characteristics

Physically Physically inactive pre

Overweight (BMI

Other Other diseases,

Hyperlipidaemia,

Peripheral arterial disease,

Hypertension,

Heart failure,

Psychiatric Psychiatric disorder,

Diabetes,

Previous angina pectoris,

Previous myocardial infarction,

Previous atrial fibrillation,

Previous transient ischaemic attack,

Previous stroke,

Risk factorsRisk for stroke

Living Living

Sex, n (%) n Sex,

Age, Age, (Q1−Q3)years, median

Demographic characteristics

Table 6: Demographics Table

value

0.117

<.0001*

0.176

0.160

1.000

0.351

<.0001*

0.036

1.000

(cont.)

)

≥ 58

, upper extremity

(95)

(2−7)

3 (3−4

(2−56)

16 (9−26)

44 (47)

15 (16)

46 (49)

91 (97)

45 (48)

(4−65)

51 (41−58)

5 (5)

2 (2)

stroke

< 58<

−

3 (3−6)

(8−100)

31 (18−43)

61 (81)

19 (25)

28 (37)

73 (97)

30 (40)

(0−57)

8 (4−18)

63 (84)

12 (16)

2 (3)

value

AD at 6 months AD post 6at months

0.228

<.0001*

0.000*

0.250

0.263

0.586

0.261

<.0001*

0.000*

1.000

5) 5)

FMA

−

≥32

−

3 (3

(2−59)

17 (12−30)

73 (56)

24 (18)

61 (46)

127 (96)

62 (47)

(0−65)

47 (21−56)

125 (95)

7 (5)

3 (3)

, Scandinavian Stroke Scale; FMA

< 32<

(86)

−

cohort at baseline, and other and cohort baseline, (n=223) at confounders

3 (3−5)

(10−100

37 (22−44)

10 (27)

13 (35)

37 (100)

13 (35)

(0−14)

4 (4−8)

27 (73)

10 (27)

1 (3)

ESS

(0−7)

3 (3−5)

(2−100)

21 (13−34)

139 (63)

49 (22)

106 (48)

217 (97)

108 (48)

(0−65)

36 (8−55)

194 (87)

29 (13)

6 (3)

Grimby Physical Level Activity Scale; SSS

−

; SGPALS, Saltin

, subscale A

and clinical characteristics of the SALGOT of the characteristics and clinical

, (%) n yes,

median (Q1−Q3)median (min−max)

, n ,(%) n

median (Q1−Q3)median (min−max)

)

(SSS≤25)

Meyer Assessment

, (Q1−Q3)median (min−max)

yes, n (%) n yes,

moderate stroke (SSS>25)

yes, n (%) n yes,

tions: BMI, Body Mass Index

confounders

via

asia

Mild

Major stroke

Thrombectomy

*: p≤0.01*:

section of of section Fugl

Abbre

predictors (baseline),

No. the days from of stroke onset to the measurement of the potential

No. hospital of days,

Other

Leg Leg paresis

Aph

Affected dominant hand

Dominant Dominant right, hand, (%) n

Affected arm, right, (%) n

FMA

Stroke severity (SSS

Table 6: Demographics Table

001*

n.a.

0.000*

95%

4; 394;

3; 2853;

n.a.

odds ratio, unadjusted; OR b

OR c

n.a.

0.000*

n.a.

0.000*

n.a.

0.000*

n.a.

0.000*

not applicable;not a OR

assessment of candidate predictors

n.a.,

the

112

95%

3; 178)3;

8; 2468;

(3; 26

n.a.

(subscale A−D);

OR b

n.a.

0.000*

n.a.

0.000*

n.a.

0.000*

n.a.

0.000*

Meyer Assessment

number days of strokefrom to onset

sample, sample, age, sex, living arrangement, previous stroke, stroke diagnosis, affected dominant leg hand,

6; 156;

7; 1417;

1; 1521;

3; 723;

95%CI

n.a.

OR a

n.a.

odds ratio adjusted for

AD≥ 32) 6 monthsAD≥ at

upper section extremity Fugl of

)

≥32

; OR ; OR c

(6 months

FMA

< 32<

arrangement

, n

stroke stroke physical activity level, number of hospital days, and

favorable outcome (FMA outcome favorable

ity, pre

for for

95% confidence interval; FMA

sample, age, livingsex,

, n

within synergies

95%CI,

, n

dds ratio

asia, stroke sever

grasp=Reference

extension

ry dysfunction=Referencery

0.01

sensory function

Partial

: Can hold : Can against a tug

: Can grasp,: Can notbut hold aagainst tug

: Cannot grasp=Reference

: Can hold : Can against a tug

: Can grasp,: Can notbut hold aagainst tug

: Cannot

: Full

: Partial

: Absent=Reference

: Full

: Partial

: Absent=Reference

: Full

: Partial

: Absent=Reference

: Full

: Absent=Reference

: Full

: Partial

: Absent=Reference

*: p≤ *:

paresis, a

odds ratio adjusted for

Abbreviations:

Full Full

Senso

Sensory function

Cylinder grasp

Pincer grasp

Finger extension

Dorsiflexion

Pronation/supination

Elbow

Shoulder abduction Candidate predictor, values Table 7: O Table

asia,

004*

0.000*

0.019

0.000*

0.026

odds adjustedratio

460

826

762

95%

6; 456 6;

3; 702 3;

5; 2 5;

2; 3 2;

8; 8 8;

3; 163;

5; 8195;

3; 673;

7; 3577;

2; 722;

7; 847;

(10

(5; 285)

(18

(1.4; 44.4)

(24

(19;740

(35

(36

(1.2; 12.4)

OR c

113

7.9

245

118

367

567

3.8

odds ratio, unadjusted; OR

0.000*

candidate predictors

; OR

)

4; 182 4;

7; 183 7;

6; 346 6;

8; 520 8;

8; 132 8;

7; 639 7;

3; 158)3;

2; 470 2;

8; 938;

95%

3; 14)3;

4; 734;

3; 333;

7; 637;

8; 448;

2; 182;

essment of

(13

(27

(25

(12

(45

(11

(31

(11

ass

(subscale A−D

the the

OR b

115

171

121

stroke

0.000*

Meyer Assessment

Fugl

5; 104 5;

9; 164 9;

3; 442 3;

2; 912;

3; 226 3;

8; 244 8;

95%CI

5; 145;

6; 516;

9; 289;

6; 726;

1; 29)1;

1; 14)1;

(12

(23;238

(4; 36

(28

(11

(23

(24

(10;71

sample, sample, age, sex, living arrangement, previous stroke, stroke diagnosis, affected dominant leghand, paresis, a

OR a

12,1

118

AD≥ 58) 6 monthsAD≥ post at

≥58

AD, AD, upper section extremity of

(6 months)

FMA

<58

FMA

odds ratio adjusted for

favorable outcome (FMA outcome favorable

for for

95% confidence interval;

, n

stroke stroke physical level,activity number of hospital days, and number of days from stroke onset to

, n

95%CI,

, n

function

dds ratio

dysfunction=Reference

Candidate predictor, values

extension,

sample, sample, sex, age, living arrangement; OR

Can grasp,Can not but hold aagainst tug

: Can hold : Can against a tug

: Cannot grasp=Reference

: Can hold : Can against a tug

: Can grasp,: Can not but hold aagainst tug

: Cannot grasp=Reference

: Full

: Partial

: Absent=Reference

: Full

: Partial

: Absent=Reference

: Full

: Partial

: Absent=Reference

: Full

: Partial

: Absent=Reference

: Full

: Partial

: Absent=Reference

*: p≤0.01*:

stroke stroke severity, pre

for

Abbreviations:

Full Full sensory

Sensory

Sensory function

Cylinder grasp

Pincer grasp

Finger extension

Dorsiflexion

Pronation/supination

Elbow

0 Shoulder abduction, n Table 8: O Table

58 Discussion

Paper I and II: Early intervention for the paretic arm after stroke (the ESS-trial)

Main results This RCT was set up to investigate the effectiveness of ESS provided during early stroke rehabilitation on the recovery of arm functioning. We demonstrated that the present version of the ESS-protocol is equivalent to arm training alone in improving hand dexterity, and reducing global disability and motor and sensory impairments in the affected arm. Both trial groups showed statistically significant improvements in all outcome measures (except pain) throughout the course of the trial. Furthermore, clinically important improvements in motor function (including hand grip strength) and global disability were found in each trial group.

Comparison with studies showing beneficial effect of ESS Despite some promising results after a single 2-hour ESS-session to the peripheral hand nerves on arm motor function and motor skills in all stroke phases (61-66), a recent systematic review with meta-analysis (40) failed to demonstrate any convincing effect of multiple ESS sessions in chronic stroke. Noteworthy, this conclusion was based on limited available data. A recently published RCT (127) delivered ESS to the peripheral hand nerves in a high intervention dose (18 sessions x 2-hour ESS followed by 4-hour task-oriented arm training) and demonstrated beneficial, long-term effect on functional capacity in subacute/chronic stroke patients with some active finger and wrist extension. Interestingly, their inactive placebo ESS had a positive, but transitory influence on the Stroke Impact Scale. In subacute stroke, one study (67) reported that 12 sessions x 20-minute task-oriented arm training preceded by 2-hour subsensory ESS to the median nerve had a positive, short-term effect. Interestingly, again, the subsensory ESS was the control intervention. Another study (128) showed that 10 sessions x 45-minute ESS provided using a stimulation glove had likewise a positive post-intervention effect. Apart from the present trial, we have identified only one study (129) conducted in the acute stroke phase. Contrary to our results, this study reported beneficial, long-term effect of ESS on hand grip strength and tip pinch strength in patients with mild-to-moderate stroke and severely impaired arm. The ESS was applied to acupuncture points, the stimulation parameters were different than ours (200µs pulses administrated at 20Hz at the maximum tolerable intensity), the ESS was not specifically followed by arm training, and the intervention dose was higher (20 sessions x 60-minute ESS) compared with our trial. Conversely, a systematic review with meta-analysis (130) demonstrated

59 benefits from ESS applied to the lower limb, particularly in acute/subacute stroke. In conclusion, these findings suggest that the available evidence on ESS delivered in the early weeks after stroke is conflicting. Moreover, whether ESS in later stroke phases is beneficial and under which conditions require further clarification.

Possible explanations for not finding a beneficial effect of ESS in the present trial In our trial, we administrated ESS in addition to task-oriented arm training, which has shown to be effective in the first 4 weeks post-stoke (48). It is possible that ESS is not sufficiently potent to facilitate improvements beyond the effect of task-oriented arm training and spontaneous neurologic recovery. Note that ESS was not combined with task-oriented training in the above- mentioned study (129) showing positive effect of ESS in the acute stroke patients. In later phases of stroke, when the contribution of spontaneous neurologic recovery to the restoration of arm functioning becomes more modest, it might be possible that ESS can play a beneficial role as suggested by single studies in subacute (67, 128) and chronic stroke (68, 127).

An insufficient intervention dose is another possible explanation for the lack of effect. We delivered a relatively low number of 12−13 ESS-sessions compared with higher numbers of 18−28 ESS-sessions in other studies demonstrating positive effect of ESS (68, 127, 129). Ideally, we would had preferred to provide the ESS-intervention throughout the whole first month post- stroke for all study participants, but financial and logistical reasons prevented us from doing this after hospital discharge. Thus, it is likely that a longer ESS-intervention period might have influenced the outcomes. However, in the trial design we addressed our concern for not achieving a sufficient intervention dose by deciding to provide the ESS-intervention during all hospitalization days (including weekend days). We cannot exclude the possibility of a placebo effect as also suggested by other authors (129), or the possibility of the placebo having a real effect.

We cannot exclude that a different ESS-protocol in terms of electrode placement and stimulation parameters may have led to other results. According to a systematic review (71) available in the design phase of our trial, there was a great variability in the stimulation protocols used so far. Our stimulation protocol was similar to the ones used in the most studies regarding key parameters such as the placement of the electrodes on the peripheral hand nerves (the novelty in the ESS-trial was adding the cutaneous stimulation to the shoulder area), pulse duration, pulse frequency, intensity of stimulation and duration of a single ESS-session. Though, the most used

60 frequency of electrical impulses was 10Hz delivered in trains at 1Hz, and our protocol deviated from this parameter with a possible unfavourable influence of the results. However, because ESS-protocols varied considerably across studies, the importance of these aspects is difficult to clarify (40, 71).

Strengths and limitations We applied outcome measures that are highly prevalent in stroke trials (mRS, FMA-UE, SSS) (107). This makes comparisons between trials possible and supports meta-analyses. The selected outcome measures captured aspects of both pure neurologic recovery (FMA-UE, PTT, and hand grip and pinch strength) and recovery of functional capacity that may partly be achieved through compensatory strategies (BBT, mRS). However, we acknowledge that clinical measures as our selected primary outcome (BBT) cannot distinguish between pure neurologic recovery and compensation. This was recently highlighted in consensus-based recommendations for measurement in stroke recovery and rehabilitation trials (131), in which there was push for the use of kinematic and kinetic measures in future intervention trials for the upper limb after stroke. All assessments were undertaken by the same blinded assessor (EG), which eliminated the risk of systematic bias due to lack of agreement between several assessors.

We employed broad eligibility criteria. All adult stroke patients living in the hospital’s catchment area, except individuals with severe pre-stroke disability, residual arm paresis from a previous stroke, contraindications to electrical stimulation, and inability to provide informed consent due to cognitive dysfunctions or limited communication skills in Danish, were eligible for enrolment in the trial. Patients with comorbidities were not excluded. Although, the final sample mostly comprised patients with mild-to-moderate stroke, moderately impaired arm function, and some finger extension. As a result, our results are generalizable only to stroke patients with these characteristics. Because the characteristics of the patients excluded from the ESS-trial (Figure 4) were not recorded, we are unable to evaluate whether those patients were different from the ones included in the trial. Few eligible patients declined to participate, and only 14% of the participants in the control group and 13% in the intervention group were lost to 6-month follow-up. Collecting follow-up data at the participants’ residences likely contributed to the limited missing data. When screening for eligible patients, we did not use any specific measure for cognitive function with a well-defined cut-off. We approached those patients who seemed to be able to provide informed consent according to the medical record and/or the

61 medical staff. Consequently, it is possible that we overlooked patients with severe strokes and severe cognitive impairments who would have been eligible for the trial.

The randomization process created two well balanced trial groups (Table 1). Yet, there was a trend that the intervention group received a higher dose of arm training during the ESS- intervention period (5.3h vs. 4h, p=0.02) and comprised less participants with severe stroke (0% vs. 8.2%, p=0.05) compared with the control group, albeit without reaching a significance level of 1%. However, both trial groups improved equally. One possible explanation is the presence of a placebo effect or a real effect of the placebo treatment. Another explanation is that the trend is too modest to have any clinical importance. Because we did not use transcranial magnetic stimulation (TMS) and MRI to assess the degree of structural neural damage, we cannot be sure that the participants with initial severe arm paresis are equally distributed between the trial groups regarding their potential for recovery (47).

The ESS-trial was not powered for subgroup analyses. Therefore, we were not able to investigate possible effects of ESS on subgroups based on, for example, severity of arm paresis, or potential for recovery based on TMS and MRI despite initial severe paresis (47).

In the design phase of the trial, we identified four other RCTs that had examined the effect of ESS on arm recovery after stroke (67-70). Among these, two trials (67, 70) used subsensory ESS as control; one trial used a completely inactive placebo, but the ESS-intervention was delivered at the participants’ homes without any risk of interaction between them, and the last trial employed a cross-over design (69). Because the ESS-trial was to be conducted in a single stroke unit with no possibility of preventing the participants from communicating with each other, we designed a placebo with a very low dose of active ESS to prevent high dropout percent in the control group. Although the low-dose ESS was unlikely to induce any real treatment effect, we cannot exclude a possible placebo effect.

Because suprasensory ESS was perceivable by nature, it was not feasible to blind the study participants to their group allocation, but they were kept unaware of which ESS-mode was the intervention and which one was the control. The ESS-staff could not be blinded either, as they adjusted the current level in the intervention group after the first 30 minutes of stimulation to prevent adaptation and maintain a suprasensory stimulation level. Due to unavailable staff

62 resources, we could not appoint another person for this task solely. Both the control and the intervention groups received the same amount of attention from the ESS-staff.

Paper III: Early predictors of the affected arm functioning at 6 months post-stroke

Main results This study evaluated the individual value of clinical tests that are easy-to-perform early after stroke for prediction of arm functioning at 6 months post-stroke. The data show the presence of partial shoulder abduction within synergies measured 3−7 days after stroke was the strongest predictor (OR c=7.3) of the ability of involving the affected arm in performance of at least basic ADL at 6 months post-stroke. Furthermore, full elbow extension within synergies (OR c=36.8), full mass finger extension (OR c=18.3), partial/full wrist dorsiflexion (OR c=19/24.5) and partial/full forearm pronation/supination (OR c=8.7/35.2) were the strongest predictors of an almost complete arm recovery 6 months post-stroke.

Comparison with other studies The data suggest that partial/full proximal arm function (shoulder abduction and elbow extension within synergies), partial/full distal arm function (forearm pronation/supination, wrist dorsiflexion, pincer and cylinder grasp, mass finger extension), and intact sensory function early after stroke are independently associated with a better arm functioning (FMA-UE-AD≥ 32/58) at 6 months post- stroke. These results are in line with findings from previous studies showing that proximal arm control (23, 25, 27, 75, 76), distal arm control (23, 74-76), and sufficiently sensory function (75-77) measured during the first month predict a better hand dexterity at 3 to ≥12 months after stroke. To the best of our knowledge, this is the first investigation that aimed to specifically identify clinical tests that predict an almost complete arm recovery after stroke (FMA-UE-AD≥58).

Interestingly, the presence of partial shoulder abduction within synergies and partial finger extension during the first week after stroke was not significantly associated with an almost complete arm recovery at 6 months post-stroke in this study, unless the participants were able to perform these movements fully (item score=2).

Strengths and limitations The advantages of using the FMA-UE-AD as outcome allows predicting improvements predominantly due to pure neurologic recovery rather than functional recovery, which includes the interference of compensatory strategies (e.g. forward-bending of the upper body to compensate for

63 impaired elbow extension in reaching tasks). Functional recovery might, however, be meaningful for patients and health care professionals since it reflects the patients’ ability to perform daily tasks relevant for their real-life. Thus, dichotomizing the FMA-UE-AD according to the proposed cut-off levels of 32 and 58 points, indicating the ability of performing at least basic ADL (92, 93) and the ability of routinely using the affected arm in ADL, respectively (94), provides an opportunity to relate the outcome to functional ability, which is a strength of this study. Furthermore, it allows us to make inferences about the functional ability of the affected arm at 6 months post-stroke based on impairments assessed during the first week after stroke. The large sample size (n=223) achieved by merging two independent cohorts of stroke patients from two hospitals in different Scandinavian countries is another strength of this study, which improves the generalizability of the results. Finally, both the predictors and the predicted outcome are clinically relevant, easy-to-perform, and suitable for the majority of the stroke patients, which makes our study valuable for clinical practice. In the present sample, the distribution of study participants with partial/full motor function at baseline was, with very few exceptions, 0 for an arm functioning of FMA-UE-AD<32 at 6 months post-stroke (see Table 7). Consequently, two limitations arised. Firstly, it was not possible to calculate the OR for a FMA-UE-AD≥32 for the majority of the motor items. Secondly, the subgroup of patients with absent motor function (a score of 0 on items of FMA-UE-AD at baseline) and sensory dysfunction (a score < 12 on FMA-UE-H at baseline) was selected as reference groups in the logistic regression analyses. It would have been of more clinical value to assess the predictive value of the selected clinical tests in the subgroups of patients with absent/partial motor function and sensory dysfunction in comparison with a reference group consisting of patients with full motor/sensory function. However, this type of data distribution is not unique to the SALGOT- ESS- sample, but rather common in acute stroke trials.

Conclusions

The key findings of the research in this PhD-thesis can be summarized as following: • ESS delivered according to the present protocol and followed by task-oriented arm training has a similar effect to task-oriented arm training alone in improving hand dexterity, motor and sensory arm functions, and in reducing global disability at 6 months post-stroke. The results pertain to patients with mild-to-moderate stroke, moderate arm impairments and some voluntary finger extension. Whether ESS is effective in early rehabilitation of stroke patients with other characteristics (e.g. severe-to-moderate stroke, potential for recovery despite initial inability of finger extension), in later time intervals after stroke, or when using a 64 different stimulation protocol (i.e. treatment dose, electrode placement, stimulation parameters) remain unanswered. • The present PhD-thesis confirms previous findings that sufficient sensory function and some proximal/distal arm movement during the first week post-stroke are associated with a better arm functioning at 6 months post-stroke in patients with moderately impaired arm motor function and mild-to-moderate stroke. We further extend those finding by identifying individual clinical tests that specifically predict an almost complete arm recovery; we demonstrated that the presence of more demanding distal arm movements (partial/full forearm pronation/supination wrist dorsiflexion, grasping ability and full finger extension) during the first week after stroke is a predictor of an almost complete recovery 6 months post-stroke in patients with same characteristics. Further studies are needed to refine clinically accessible prediction models for patients who do not show any initial volitional movement early after stroke.

Future considerations

Important questions are still unanswered and should be addressed in future research: • What are the most effective therapeutic interventions (i.e. methods and treatment doses) for recovery of arm functioning after stroke? Probably, different methods and treatment doses are needed in different time intervals after stroke and in patients with different characteristics. Future trials should design and investigate novel therapeutic interventions based on an understanding of the underlying neurophysiological recovery mechanisms to maximize the success rate of the trial and to use the limited research funding rationally. Concerning the acute stroke phase, the challenge seems to be designing interventions that are able to facilitate motor and sensory function beyond the level of spontaneous neurologic recovery. Furthermore, we encourage the development and the use of outcome measures that cover all the domains of human functioning (i.e. body functions, activities and participation) and are particularly responsive to change (e.g. kinetic and kinematic measures). Very few studies have included measures of the actual activity performance in the real-life and measures of participation, although arguably these are the most meaningful domains for the patients. • What are the most accurate predictors (i.e. individual predictors or prediction models) of future arm functioning that can easily be performed/implemented in the early clinical

65 examinations after stroke? How long in the future can we accurately predict arm functioning using these early predictors?

Acknowledgements

I would like to express my deepest appreciation and thanks to all people and institutions that have supported this work: • my supervisors, S. Peter Magnusson, professor, Hanne Christensen, professor, and Christian Couppé, PhD, for inspiring discussions and valuable personal and professional guidance; I could not imagine having better supervisors and mentors for entering the field of research. • my collaborators/co-authors, Volkert Siersma, statistician, PhD, Metter Søndergaard, MD, Katharina S. Sunnerhagen, professor, Hanna C. Persson, PhD, and Margit Alt Murphy, PhD, for valuable discussions and collaborations. • professor Katharina S. Sunnerhagen and the research group at the Department of Clinical Neuroscience and Rehabilitation, Institute of Neuroscience and Physiology, Sahlgrenska Academy, University of Gothenburg, Sweden, for giving me the opportunity to follow their work during my PhD-student visit in February-March 2018, and for introducing me to Swedish language, culture, and the city of Gothenburg. • all trial participants • the Department of Physical and Occupational Therapy, BFH • all the therapists at the rehabilitation stroke unit, BFH, for integrating the ESS-trial in their clinical work, and for providing and recording the arm training according to the trial protocol. • the health care staff responsible for ESS-treatment and arm training during weekends for their enormous flexibility and professional engagement. • the nursing staff at the rehabilitation stroke unit, BFH, for integrating the ESS-trial procedures in the daily clinical practice, and for being supportive in the enrolment of trial participants. • all the collaborating rehabilitation centres in the community (Hjerneskade- og Rehabiliteringscentret, Rehabiliteringscenter Emdrup, Træningscenter Bispebjerg, Aktivitetscentret Bispebjerg, Træningscenter Indre By/Østerbro, Rehabiliteringscenter Indre By/Østerbro, Sundhedscentret Stockflethsvej, Frederiksberg Rehabilitering Lioba, Frederiksbergs Kommunes Rehabiliteringsenhed Valby, Center for Neurorehabilitering

66 Kurhus, Center for Sundhed og Omsorg/Sundhed og Træning Korsør, Sundhed og Genoptræning Amager, and Center for Rehabilitation of Brain Injury) and nursing homes (Ingeborggården, Søndervang, Poppelbo) for recording the training provided to the trial participants. A special thank goes to our contact persons at each of the above-mentioned institutions. • the administrative personnel at BFH for taking care of the group allocation of the participants in the ESS-trial. • the Department of Physical and Occupational Therapy at Rigshospitalet/Glostrup Hospital for introducing the trial therapists to task-oriented arm training. • all grant providers for their generous financial support: the Capital Region of Denmark, Foundation for Health Research; Bevica Fonden; Lundbeck Foundation [FP 68/2013]; the Danish Association of Occupational Therapists [FF 1 14-3]; Direktør Jacob Madsen & hustru Olga Madsen’s fond [5507], and the Department of Physical and Occupational Therapy, BFH. • my family and friends for their continuous encouragement throughout the whole PhD- process, particularly to my 16-year old son, who has shown so much understanding for having a parent that spent many evening and weekend hours studying during the last years.

References

1. Ghaziani E, Couppe C, Henkel C, Siersma V, Sondergaard M, Christensen H, et al. Electrical somatosensory stimulation followed by motor training of the paretic upper limb in acute stroke: study protocol for a randomized controlled trial. Trials. 2017;18(1):84. 2. World Health Organization. Stroke, Cerebrovascular accident: World Health Organization; [Available from: http://www.who.int/topics/cerebrovascular_accident/en/. 3. Camak DJ. Addressing the burden of stroke caregivers: a literature review. Journal of clinical nursing. 2015;24(17-18):2376-82. 4. Feigin VL, Barker-Collo S, McNaughton H, Brown P, Kerse N. Long-term neuropsychological and functional outcomes in stroke survivors: current evidence and perspectives for new research. International journal of stroke : official journal of the International Stroke Society. 2008;3(1):33-40. 5. Jennum P, Iversen HK, Ibsen R, Kjellberg J. Cost of stroke: a controlled national study evaluating societal effects on patients and their partners. BMC health services research. 2015;15:466. 6. O'Donnell MJ, Xavier D, Liu L, Zhang H, Chin SL, Rao-Melacini P, et al. Risk factors for ischaemic and intracerebral haemorrhagic stroke in 22 countries (the INTERSTROKE study): a case-control study. Lancet (London, England). 2010;376(9735):112-23. 7. Murray CJ, Lopez AD. Mortality by cause for eight regions of the world: Global Burden of Disease Study. Lancet (London, England). 1997;349(9061):1269-76.

67 8. Feigin VL, Norrving B, Mensah GA. Global Burden of Stroke. Circ Res. 2017;120(3):439-48. 9. Feigin VL, Krishnamurthi RV, Parmar P, Norrving B, Mensah GA, Bennett DA, et al. Update on the Global Burden of Ischemic and Hemorrhagic Stroke in 1990-2013: The GBD 2013 Study. Neuroepidemiology. 2015;45(3):161-76. 10. Krishnamurthi RV, Moran AE, Feigin VL, Barker-Collo S, Norrving B, Mensah GA, et al. Stroke Prevalence, Mortality and Disability-Adjusted Life Years in Adults Aged 20-64 Years in 1990-2013: Data from the Global Burden of Disease 2013 Study. Neuroepidemiology. 2015;45(3):190-202. 11. Goeggel Simonetti B, Mono ML, Huynh-Do U, Michel P, Odier C, Sztajzel R, et al. Risk factors, aetiology and outcome of ischaemic stroke in young adults: the Swiss Young Stroke Study (SYSS). J Neurol. 2015;262(9):2025-32. 12. Maaijwee NA, Rutten-Jacobs LC, Schaapsmeerders P, van Dijk EJ, de Leeuw FE. Ischaemic stroke in young adults: risk factors and long-term consequences. Nat Rev Neurol. 2014;10(6):315-25. 13. Danish Health Authority. Sygdomsbyrden i Danmark: Sygdomme. 2015. 14. World Health Organization. International classification of functioning, disability and health : ICF. 15. Rand D, Eng JJ. Predicting daily use of the affected upper extremity 1 year after stroke. Journal of stroke and cerebrovascular diseases : the official journal of National Stroke Association. 2015;24(2):274-83. 16. Cortes JC, Goldsmith J, Harran MD, Xu J, Kim N, Schambra HM, et al. A Short and Distinct Time Window for Recovery of Arm Motor Control Early After Stroke Revealed With a Global Measure of Trajectory Kinematics. Neurorehabilitation and neural repair. 2017;31(6):552-60. 17. Nichols-Larsen DS, Clark PC, Zeringue A, Greenspan A, Blanton S. Factors influencing stroke survivors' quality of life during subacute recovery. Stroke. 2005;36(7):1480-4. 18. Lawrence ES, Coshall C, Dundas R, Stewart J, Rudd AG, Howard R, et al. Estimates of the prevalence of acute stroke impairments and disability in a multiethnic population. Stroke. 2001;32(6):1279-84. 19. Nakayama H, Jorgensen HS, Raaschou HO, Olsen TS. Recovery of upper extremity function in stroke patients: the Copenhagen Stroke Study. Archives of physical medicine and rehabilitation. 1994;75(4):394-8. 20. Persson HC, Parziali M, Danielsson A, Sunnerhagen KS. Outcome and upper extremity function within 72 hours after first occasion of stroke in an unselected population at a stroke unit. A part of the SALGOT study. BMC neurology. 2012;12:162. 21. Rathore SS, Hinn AR, Cooper LS, Tyroler HA, Rosamond WD. Characterization of incident stroke signs and symptoms: findings from the atherosclerosis risk in communities study. Stroke. 2002;33(11):2718-21. 22. Fleming MK, Newham DJ, Roberts-Lewis SF, Sorinola IO. Self-perceived utilization of the paretic arm in chronic stroke requires high upper limb functional ability. Archives of physical medicine and rehabilitation. 2014;95(5):918-24. 23. Kong KH, Lee J. Temporal recovery and predictors of upper limb dexterity in the first year of stroke: a prospective study of patients admitted to a rehabilitation centre. NeuroRehabilitation. 2013;32(2):345-50. 24. Kwakkel G, Kollen BJ, van der Grond J, Prevo AJ. Probability of regaining dexterity in the flaccid upper limb: impact of severity of paresis and time since onset in acute stroke. Stroke. 2003;34(9):2181-6. 25. Nijland RH, van Wegen EE, Harmeling-van der Wel BC, Kwakkel G. Presence of finger extension and shoulder abduction within 72 hours after stroke predicts functional

68 recovery: early prediction of functional outcome after stroke: the EPOS cohort study. Stroke. 2010;41(4):745-50. 26. Michielsen ME, Selles RW, Stam HJ, Ribbers GM, Bussmann JB. Quantifying nonuse in chronic stroke patients: a study into paretic, nonparetic, and bimanual upper-limb use in daily life. Archives of physical medicine and rehabilitation. 2012;93(11):1975-81. 27. Houwink A, Nijland RH, Geurts AC, Kwakkel G. Functional recovery of the paretic upper limb after stroke: who regains hand capacity? Archives of physical medicine and rehabilitation. 2013;94(5):839-44. 28. Murphy TH, Corbett D. Plasticity during stroke recovery: from synapse to behaviour. Nature reviews Neuroscience. 2009;10(12):861-72. 29. Kwakkel G. Impact of intensity of practice after stroke: issues for consideration. Disability and rehabilitation. 2006;28(13-14):823-30. 30. Kwakkel G, Kollen BJ. Predicting activities after stroke: what is clinically relevant? International journal of stroke : official journal of the International Stroke Society. 2013;8(1):25-32. 31. Lee KB, Lim SH, Kim KH, Kim KJ, Kim YR, Chang WN, et al. Six-month functional recovery of stroke patients: a multi-time-point study. International journal of rehabilitation research Internationale Zeitschrift fur Rehabilitationsforschung Revue internationale de recherches de readaptation. 2015;38(2):173-80. 32. van Kordelaar J, van Wegen E, Kwakkel G. Impact of time on quality of motor control of the paretic upper limb after stroke. Archives of physical medicine and rehabilitation. 2014;95(2):338-44. 33. Verheyden G, Nieuwboer A, De Wit L, Thijs V, Dobbelaere J, Devos H, et al. Time course of trunk, arm, leg, and functional recovery after ischemic stroke. Neurorehabilitation and neural repair. 2008;22(2):173-9. 34. Winter J, Hunter S, Sim J, Crome P. Hands-on therapy interventions for upper limb motor dysfunction following stroke (Cochrane review) [with consumer summary]. Cochrane Database of Systematic Reviews 2011;Issue 6:Epub. 2011. 35. Harris JE, Eng JJ. Strength training improves upper-limb function in individuals with stroke: a meta-analysis. Stroke. 2010;41(1):136-40. 36. Ada L, Foongchomcheay A, Canning C. Supportive devices for preventing and treating subluxation of the shoulder after stroke (Cochrane review) [with consumer summary]. Cochrane Database of Systematic Reviews 2005;Issue 1. 2005. 37. Lannin NA, Herbert RD. Is hand splinting effective for adults following stroke? A systematic review and methodologic critique of published research. Clinical rehabilitation. 2003;17(8):807-16. 38. Tyson SF, Kent RM. The effect of upper limb orthotics after stroke: a systematic review. Neurorehabilitation 2011;28(1):29-36. 2011. 39. Zeng W, Guo Y, Wu G, Liu X, Fang Q. Mirror therapy for motor function of the upper extremity in patients with stroke: A meta-analysis. Journal of rehabilitation medicine. 2018;50(1):8-15. 40. Grant VM, Gibson A, Shields N. Somatosensory stimulation to improve hand and upper limb function after stroke-a systematic review with meta-analyses. Topics in stroke rehabilitation. 2017:1-11. 41. French B, Thomas LH, Coupe J, McMahon NE, Connell L, Harrison J, et al. Repetitive task training for improving functional ability after stroke. Cochrane Database Syst Rev. 2016;11:CD006073. 42. Mehrholz J, Pohl M, Platz T, Kugler J, Elsner B. Electromechanical and robot- assisted arm training for improving activities of daily living, arm function, and arm muscle strength after stroke. Cochrane Database Syst Rev. 2015(11):Cd006876.

69 43. Saposnik G, Levin M, Outcome Research Canada Working G. Virtual reality in stroke rehabilitation: a meta-analysis and implications for clinicians. Stroke 2011 May;42(5):1380-1386. 2011. 44. Corbetta D, Sirtori V, Castellini G, Moja L, Gatti R. Constraint-induced movement therapy for upper extremities in people with stroke. Cochrane Database Syst Rev. 2015(10):CD004433. 45. Popovic DB, Sinkaer T, Popovic MB. Electrical stimulation as a means for achieving recovery of function in stroke patients. NeuroRehabilitation. 2009;25(1):45-58. 46. de Kroon JR, van der Lee JH, MJ IJ, Lankhorst GJ. Therapeutic electrical stimulation to improve motor control and functional abilities of the upper extremity after stroke: a systematic review. Clinical rehabilitation. 2002;16(4):350-60. 47. Stinear CM, Byblow WD, Ackerley SJ, Smith MC, Borges VM, Barber PA. PREP2: A biomarker-based algorithm for predicting upper limb function after stroke. Ann Clin Transl Neurol. 2017;4(11):811-20. 48. Wattchow KA, McDonnell MN, Hillier SL. Rehabilitation Interventions for Upper Limb Function in the First Four Weeks Following Stroke: A Systematic Review and Meta- Analysis of the Evidence. Archives of physical medicine and rehabilitation. 2017. 49. INVOLVE. Top ten research priorities relating after stroke: National Institute for Health Research; [Available from: http://www.invo.org.uk/top-ten-research-priorities-relating- after-stroke/. 50. Han L, Law-Gibson D, Reding M. Key neurological impairments influence function-related group outcomes after stroke. Stroke. 2002;33(7):1920-4. 51. Nudo RJ, Friel KM, Delia SW. Role of sensory deficits in motor impairments after injury to primary motor cortex. Neuropharmacology. 2000;39(5):733-42. 52. Dobkin BH. Do electrically stimulated sensory inputs and movements lead to long- term plasticity and rehabilitation gains? Current opinion in neurology. 2003;16(6):685-91. 53. Nowak DA, Grefkes C, Ameli M, Fink GR. Interhemispheric competition after stroke: brain stimulation to enhance recovery of function of the affected hand. Neurorehabilitation and neural repair. 2009;23(7):641-56. 54. Golaszewski SM, Bergmann J, Christova M, Nardone R, Kronbichler M, Rafolt D, et al. Increased motor cortical excitability after whole-hand electrical stimulation: a TMS study. Clinical neurophysiology : official journal of the International Federation of Clinical Neurophysiology. 2010;121(2):248-54. 55. Golaszewski SM, Siedentopf CM, Koppelstaetter F, Rhomberg P, Guendisch GM, Schlager A, et al. Modulatory effects on human sensorimotor cortex by whole-hand afferent electrical stimulation. Neurology. 2004;62(12):2262-9. 56. Kaelin-Lang A, Luft AR, Sawaki L, Burstein AH, Sohn YH, Cohen LG. Modulation of human corticomotor excitability by somatosensory input. The Journal of physiology. 2002;540(Pt 2):623-33. 57. McKay D, Brooker R, Giacomin P, Ridding M, Miles T. Time course of induction of increased human motor cortex excitability by nerve stimulation. Neuroreport. 2002;13(10):1271-3. 58. Ridding MC, McKay DR, Thompson PD, Miles TS. Changes in corticomotor representations induced by prolonged peripheral nerve stimulation in humans. Clinical neurophysiology : official journal of the International Federation of Clinical Neurophysiology. 2001;112(8):1461-9. 59. Tinazzi M, Zarattini S, Valeriani M, Romito S, Farina S, Moretto G, et al. Long- lasting modulation of human motor cortex following prolonged transcutaneous electrical nerve stimulation (TENS) of forearm muscles: evidence of reciprocal inhibition and facilitation. Experimental brain research. 2005;161(4):457-64.

70 60. Wu CW, van Gelderen P, Hanakawa T, Yaseen Z, Cohen LG. Enduring representational plasticity after somatosensory stimulation. NeuroImage. 2005;27(4):872-84. 61. Celnik P, Hummel F, Harris-Love M, Wolk R, Cohen LG. Somatosensory stimulation enhances the effects of training functional hand tasks in patients with chronic stroke. Archives of physical medicine and rehabilitation. 2007;88(11):1369-76. 62. Conforto AB, Cohen LG, dos Santos RL, Scaff M, Marie SK. Effects of somatosensory stimulation on motor function in chronic cortico-subcortical strokes. J Neurol. 2007;254(3):333-9. 63. Conforto AB, Kaelin-Lang A, Cohen LG. Increase in hand muscle strength of stroke patients after somatosensory stimulation. Annals of neurology. 2002;51(1):122-5. 64. Klaiput A, Kitisomprayoonkul W. Increased pinch strength in acute and subacute stroke patients after simultaneous median and ulnar sensory stimulation. Neurorehabilitation and neural repair. 2009;23(4):351-6. 65. Koesler IB, Dafotakis M, Ameli M, Fink GR, Nowak DA. Electrical somatosensory stimulation improves movement kinematics of the affected hand following stroke. Journal of neurology, neurosurgery, and psychiatry. 2009;80(6):614-9. 66. Sawaki L, Wu CW, Kaelin-Lang A, Cohen LG. Effects of somatosensory stimulation on use-dependent plasticity in chronic stroke. Stroke. 2006;37(1):246-7. 67. Conforto AB, Ferreiro KN, Tomasi C, dos Santos RL, Moreira VL, Marie SK, et al. Effects of somatosensory stimulation on motor function after subacute stroke. Neurorehabilitation and neural repair. 2010;24(3):263-72. 68. Dos Santos-Fontes RL, Ferreiro de Andrade KN, Sterr A, Conforto AB. Home- based nerve stimulation to enhance effects of motor training in patients in the chronic phase after stroke: a proof-of-principle study. Neurorehabilitation and neural repair. 2013;27(6):483-90. 69. Ikuno K, Kawaguchi S, Kitabeppu S, Kitaura M, Tokuhisa K, Morimoto S, et al. Effects of peripheral sensory nerve stimulation plus task-oriented training on upper extremity function in patients with subacute stroke: a pilot randomized crossover trial. Clinical rehabilitation. 2012;26(11):999-1009. 70. Sullivan JE, Hurley D, Hedman LD. Afferent stimulation provided by glove electrode during task-specific arm exercise following stroke. Clinical rehabilitation. 2012;26(11):1010-20. 71. Laufer Y, Elboim-Gabyzon M. Does sensory transcutaneous electrical stimulation enhance motor recovery following a stroke? A systematic review. Neurorehabilitation and neural repair. 2011;25(9):799-809. 72. Smania N, Paolucci S, Tinazzi M, Borghero A, Manganotti P, Fiaschi A, et al. Active finger extension: a simple movement predicting recovery of arm function in patients with acute stroke. Stroke; a journal of cerebral circulation. 2007;38(3):1088-90. 73. Snickars J, Persson HC, Sunnerhagen KS. Early clinical predictors of motor function in the upper extremity one month post-stroke. J Rehabil Med. 2017;49(3):216-22. 74. Higgins J, Mayo NE, Desrosiers J, Salbach NM, Ahmed S. Upper-limb function and recovery in the acute phase poststroke. Journal of rehabilitation research and development. 2005;42(1):65-76. 75. Kong KH, Chua KS, Lee J. Recovery of upper limb dexterity in patients more than 1 year after stroke: Frequency, clinical correlates and predictors. NeuroRehabilitation. 2011;28(2):105-11. 76. Au-Yeung SS, Hui-Chan CW. Predicting recovery of dextrous hand function in acute stroke. Disability and rehabilitation. 2009;31(5):394-401. 77. Winters C, Kwakkel G, Nijland R, van Wegen E. When Does Return of Voluntary Finger Extension Occur Post-Stroke? A Prospective Cohort Study. PloS one. 2016;11(8):e0160528.

71 78. Coupar F, Pollock A, Rowe P, Weir C, Langhorne P. Predictors of upper limb recovery after stroke: a systematic review and meta-analysis. Clinical rehabilitation. 2012;26(4):291-313. 79. Kwah LK, Herbert RD. Prediction of Walking and Arm Recovery after Stroke: A Critical Review. Brain Sci. 2016;6(4). 80. Harvey RL. Predictors of Functional Outcome Following Stroke. Phys Med Rehabil Clin N Am. 2015;26(4):583-98. 81. Alt Murphy M, Persson HC, Danielsson A, Broeren J, Lundgren-Nilsson A, Sunnerhagen KS. SALGOT--Stroke Arm Longitudinal study at the University of Gothenburg, prospective cohort study protocol. BMC neurology. 2011;11:56. 82. Bustren EL, Sunnerhagen KS, Alt Murphy M. Movement Kinematics of the Ipsilesional Upper Extremity in Persons With Moderate or Mild Stroke. Neurorehabilitation and neural repair. 2017;31(4):376-86. 83. Watson Te. Electrotherapy. Evidence-based practice. 12. ed: Elsevier; 2008. 84. Mathiowetz V, Volland G, Kashman N, Weber K. Adult norms for the Box and Block Test of manual dexterity. The American journal of occupational therapy : official publication of the American Occupational Therapy Association. 1985;39(6):386-91. 85. Santisteban L, Teremetz M, Bleton JP, Baron JC, Maier MA, Lindberg PG. Upper Limb Outcome Measures Used in Stroke Rehabilitation Studies: A Systematic Literature Review. PloS one. 2016;11(5):e0154792. 86. Chen HM, Chen CC, Hsueh IP, Huang SL, Hsieh CL. Test-retest reproducibility and smallest real difference of 5 hand function tests in patients with stroke. Neurorehabilitation and neural repair. 2009;23(5):435-40. 87. Platz T, Pinkowski C, van Wijck F, Kim IH, di Bella P, Johnson G. Reliability and validity of arm function assessment with standardized guidelines for the Fugl-Meyer Test, Action Research Arm Test and Box and Block Test: a multicentre study. Clinical rehabilitation. 2005;19(4):404-11. 88. Lin KC, Chuang LL, Wu CY, Hsieh YW, Chang WY. Responsiveness and validity of three dexterous function measures in stroke rehabilitation. Journal of rehabilitation research and development. 2010;47(6):563-71. 89. The Box and Block Test The Rehabilitation Measure Database [updated April 26, 2012. Available from: https://www.sralab.org/rehabilitation-measures/box-and-block-test. 90. Hjermstad MJ, Fayers PM, Haugen DF, Caraceni A, Hanks GW, Loge JH, et al. Studies comparing Numerical Rating Scales, Verbal Rating Scales, and Visual Analogue Scales for assessment of pain intensity in adults: a systematic literature review. J Pain Symptom Manage. 2011;41(6):1073-93. 91. Fugl-Meyer AR, Jaasko L, Leyman I, Olsson S, Steglind S. The post-stroke hemiplegic patient. 1. a method for evaluation of physical performance. Scand J Rehabil Med. 1975;7(1):13-31. 92. Alt Murphy M, Willen C, Sunnerhagen KS. Kinematic variables quantifying upper- extremity performance after stroke during reaching and drinking from a glass. Neurorehabilitation and neural repair. 2011;25(1):71-80. 93. Persson HC, Alt Murphy M, Danielsson A, Lundgren-Nilsson A, Sunnerhagen KS. A cohort study investigating a simple, early assessment to predict upper extremity function after stroke - a part of the SALGOT study. BMC neurology. 2015;15:92. 94. Pang MY, Harris JE, Eng JJ. A community-based upper-extremity group exercise program improves motor function and performance of functional activities in chronic stroke: a randomized controlled trial. Archives of physical medicine and rehabilitation. 2006;87(1):1-9. 95. Alt Murphy M, Resteghini C, Feys P, Lamers I. An overview of systematic reviews on upper extremity outcome measures after stroke. BMC neurology. 2015;15:29.

72 96. Gladstone DJ, Danells CJ, Black SE. The fugl-meyer assessment of motor recovery after stroke: a critical review of its measurement properties. Neurorehabilitation and neural repair. 2002;16(3):232-40. 97. Page SJ, Fulk GD, Boyne P. Clinically important differences for the upper- extremity Fugl-Meyer Scale in people with minimal to moderate impairment due to chronic stroke. Physical therapy. 2012;92(6):791-8. 98. Arya KN, Verma R, Garg RK. Estimating the minimal clinically important difference of an upper extremity recovery measure in subacute stroke patients. Topics in stroke rehabilitation. 2011;18 Suppl 1:599-610. 99. Barak S, Duncan PW. Issues in selecting outcome measures to assess functional recovery after stroke. NeuroRx. 2006;3(4):505-24. 100. Sullivan KJ, Tilson JK, Cen SY, Rose DK, Hershberg J, Correa A, et al. Fugl- Meyer assessment of sensorimotor function after stroke: standardized training procedure for clinical practice and clinical trials. Stroke. 2011;42(2):427-32. 101. Fugl-Meyer AR JL, Leyman I, Olsson S, Steglind S. Fugl-Meyer Assessment - Upper Extremity (FMA-UE) University of Gotheburg. Institute of Neuroscience and Physiology1975 [updated 11 November 2015. Available from: https://neurophys.gu.se/english/departments/clinical_neuroscience_and_rehabilitation/rehabilitati on-medicine/fugl-meyer. 102. Lang CE, Edwards DF, Birkenmeier RL, Dromerick AW. Estimating minimal clinically important differences of upper-extremity measures early after stroke. Archives of physical medicine and rehabilitation. 2008;89(9):1693-700. 103. Sang CN, Max MB, Gracely RH. Stability and reliability of detection thresholds for human A-Beta and A-delta sensory afferents determined by cutaneous electrical stimulation. J Pain Symptom Manage. 2003;25(1):64-73. 104. Eek E, Engardt M. Assessment of the perceptual threshold of touch (PTT) with high-frequency transcutaneous electric nerve stimulation (Hf/TENS) in elderly patients with stroke: a reliability study. Clinical rehabilitation. 2003;17(8):825-34. 105. Eek E, Holmqvist LW, Sommerfeld DK. Adult norms of the perceptual threshold of touch (PTT) in the hands and feet in relation to age, gender, and right and left side using transcutaneous electrical nerve stimulation. Physiother Theory Pract. 2012;28(5):373-83. 106. Rankin J. Cerebral vascular accidents in patients over the age of 60. II. Prognosis. Scott Med J. 1957;2(5):200-15. 107. Quinn TJ, Dawson J, Walters MR, Lees KR. Functional outcome measures in contemporary stroke trials. International journal of stroke : official journal of the International Stroke Society. 2009;4(3):200-5. 108. Quinn TJ, Dawson J, Walters MR, Lees KR. Reliability of the modified Rankin Scale: a systematic review. Stroke. 2009;40(10):3393-5. 109. Wilson JT, Hareendran A, Grant M, Baird T, Schulz UG, Muir KW, et al. Improving the assessment of outcomes in stroke: use of a structured interview to assign grades on the modified Rankin Scale. Stroke. 2002;33(9):2243-6. 110. Wilson JT, Hareendran A, Hendry A, Potter J, Bone I, Muir KW. Reliability of the modified Rankin Scale across multiple raters: benefits of a structured interview. Stroke. 2005;36(4):777-81. 111. Bruno A, Shah N, Lin C, Close B, Hess DC, Davis K, et al. Improving modified Rankin Scale assessment with a simplified questionnaire. Stroke. 2010;41(5):1048-50. 112. Modified Rankin Scale Training and Certification [Available from: https://secure.trainingcampus.net/uas/modules/trees/windex.aspx?rx=rankin- asa.trainingcampus.net&

73 113. Dromerick AW, Edwards DF, Diringer MN. Sensitivity to changes in disability after stroke: a comparison of four scales useful in clinical trials. Journal of rehabilitation research and development. 2003;40(1):1-8. 114. Multicenter trial of hemodilution in ischemic stroke--background and study protocol. Scandinavian Stroke Study Group. Stroke. 1985;16(5):885-90. 115. Boysen G, Christensen H. Stroke severity determines body temperature in acute stroke. Stroke. 2001;32(2):413-7. 116. Grimby G, Borjesson M, Jonsdottir IH, Schnohr P, Thelle DS, Saltin B. The "Saltin-Grimby Physical Activity Level Scale" and its application to health research. Scand J Med Sci Sports. 2015;25 Suppl 4:119-25. 117. Saltin B, Grimby G. Physiological analysis of middle-aged and old former athletes. Comparison with still active athletes of the same ages. Circulation. 1968;38(6):1104-15. 118. Brott T, Adams HP, Jr., Olinger CP, Marler JR, Barsan WG, Biller J, et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke. 1989;20(7):864- 70. 119. Grimby G. Physical activity and muscle training in the elderly. Acta Med Scand Suppl. 1986;711:233-7. 120. Mattiasson-Nilo I, Sonn U, Johannesson K, Gosman-Hedstrom G, Persson GB, Grimby G. Domestic activities and walking in the elderly: evaluation from a 30-hour heart rate recording. Aging (Milano). 1990;2(2):191-8. 121. Gray LJ, Ali M, Lyden PD, Bath PM, Virtual International Stroke Trials Archive C. Interconversion of the National Institutes of Health Stroke Scale and Scandinavian Stroke Scale in acute stroke. J Stroke Cerebrovasc Dis. 2009;18(6):466-8. 122. Washburn RA, Smith KW, Jette AM, Janney CA. The Physical Activity Scale for the Elderly (PASE): development and evaluation. J Clin Epidemiol. 1993;46(2):153-62. 123. Hedblad B, Jonsson S, Nilsson P, Engstrom G, Berglund G, Janzon L. Obesity and myocardial infarction--vulnerability related to occupational level and marital status. A 23-year follow-up of an urban male Swedish population. Journal of internal medicine. 2002;252(6):542- 50. 124. Olsen TS. Arm and leg paresis as outcome predictors in stroke rehabilitation. Stroke; a journal of cerebral circulation. 1990;21(2):247-51. 125. WHO. Waist Circumference and Waist–Hip Ratio: Report of a WHO Expert Consultation 2008 [Available from: http://apps.who.int/iris/bitstream/10665/44583/1/9789241501491_eng.pdf. 126. Adams HP, Jr., Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke. 1993;24(1):35-41. 127. Carrico C, Westgate PM, Powell ES, Chelette KC, Nichols L, Pettigrew LC, et al. Nerve Stimulation Enhances Task-Oriented Training for Moderate-to-Severe Hemiparesis 3-12 Months after Stroke: A Randomized Trial. American journal of physical medicine & rehabilitation. 2018. 128. Kattenstroth JC, Kalisch T, Sczesny-Kaiser M, Greulich W, Tegenthoff M, Dinse HR. Daily repetitive sensory stimulation of the paretic hand for the treatment of sensorimotor deficits in patients with subacute stroke: RESET, a randomized, sham-controlled trial. BMC neurology. 2018;18(1):2. 129. Au-Yeung SS, Hui-Chan CW. Electrical acupoint stimulation of the affected arm in acute stroke: a placebo-controlled randomized clinical trial. Clinical rehabilitation. 2014;28(2):149-58.

74 130. Kwong PW, Ng GY, Chung RC, Ng SS. Transcutaneous electrical nerve stimulation improves walking capacity and reduces spasticity in stroke survivors: a systematic review and meta-analysis. Clinical rehabilitation. 2017:269215517745349. 131. Kwakkel G, Lannin NA, Borschmann K, English C, Ali M, Churilov L, et al. Standardized measurement of sensorimotor recovery in stroke trials: Consensus-based core recommendations from the Stroke Recovery and Rehabilitation Roundtable. International journal of stroke : official journal of the International Stroke Society. 2017;12(5):451-61.

Appendix (Paper I, Paper II, Paper III)

76 Paper I

77 Ghaziani et al. Trials (2017) 18:84 DOI 10.1186/s13063-017-1815-9

STUDY PROTOCOL Open Access Electrical somatosensory stimulation followed by motor training of the paretic upper limb in acute stroke: study protocol for a randomized controlled trial Emma Ghaziani1,2* , Christian Couppé1,3,4, Cecilie Henkel2, Volkert Siersma5, Mette Søndergaard6, Hanne Christensen2,6 and S. Peter Magnusson1,3,4

Abstract Background: Upper limb paresis is one of the most frequent and persistent impairments following stroke. Only 12– 34% of stroke patients achieve full recovery of upper limb functioning, which seems to be required to habitually use the affected arm in daily tasks. Although the recovery of upper limb functioning is most pronounced during the first 4 weeks post stroke, there are few studies investigating the effect of rehabilitation during this critical time window. The purpose of this trial is to determine the effect of electrical somatosensory stimulation (ESS) initiated in the acute stroke phase on the recovery of upper limb functioning in a nonselected sample of stroke patients. Methods/design: A sample of 102 patients with upper limb paresis of varying degrees of severity is assigned to either the intervention or the control group using stratified random sampling. The intervention group receives ESS plus usual rehabilitation and the control group receives sham ESS plus usual rehabilitation. The intervention is applied as 1 h of ESS/sham ESS daily, followed by motor training of the affected upper limb. The ESS/sham ESS treatment is initiated within 7 days from stroke onset and it is delivered during hospitalization, but no longer than 4 weeks post stroke. The primary outcome is hand dexterity assessed by the Box and Block Test; secondary outcomes are the Fugl-Meyer Assessment, hand grip strength, pinch strength, perceptual threshold of touch, degree of pain, and modified Rankin Scale score. Outcome measurements are conducted at baseline, post intervention and at 6-month follow-up. Discussion: Because of the wide inclusion criteria, we believe that the results can be generalized to the larger population of patients with a first-ever stroke who present with an upper limb paresis of varying severity. On the other hand, the sample size (n = 102) may preclude subgroup analyses in such a heterogeneous sample. The sham ESS treatment totals a mere 2% of the active ESS treatment delivered to the intervention group per ESS session, and we consider that this dose is too small to induce a treatment effect. Trial registration: ClinicalTrials.gov, NCT02250365. Registered on 18 September 2014. Keywords: Acute stroke, Upper extremity, Recovery of function, Electrical stimulation, Rehabilitation, Longitudinal studies

* Correspondence: [email protected] 1Department of Physical and Occupational Therapy, Bispebjerg Hospital, Bispebjerg Bakke 23, bldg. 10, 2400, Copenhagen, Denmark 2Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark Full list of author information is available at the end of the article

© The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Ghaziani et al. Trials (2017) 18:84 Page 2 of 9

Background training, might improve motor skills of the paretic upper Stroke is ranked as the third largest cause of disease bur- limb in subacute [32, 33] and chronic stroke patients den globally [1], causing substantial physical, psycho- [34], and, moreover, that these positive results seems to logical and financial demands on patients, families, and be long lasting [34]. However, the effect of ESS in con- societies at large [2–4]. Upper limb paresis is one of the junction with motor training has never been investigated most frequent impairments following stroke and affects in acute stroke patients. It is noteworthy that ESS is be- 48–77% of patients in the acute stroke phase [5–7]. nign in nature, causes patients minimal discomfort and Moreover, upper limb paresis has been identified as a adverse effects (itch and blushing), is relatively inexpen- major obstacle to regaining independence in activities of sive and can easily be incorporated into clinical practice daily living (ADLs) [8]. In fact, only 12–34% of the pa- [35]. Therefore, it would be valuable to establish the ef- tients achieve full functional recovery of the affected fect of multiple sessions of ESS in conjunction with upper limb at 6 months post stroke [9, 10]. This repre- motor training in the restoration of upper limb function- sents a considerable challenge since near complete func- ing in the acute stroke phase. tional recovery is required to routinely involve the The purpose of the present trial is to investigate the affected upper limb in performing ADLs [11]. effect of multiple sessions of ESS treatment accompan- Recovery of upper limb functioning is typically pro- ied by motor training on the recovery of the affected nounced during the first month and subsequently levels upper limb following stroke. The ESS treatment is initi- off by 6 months post stroke [12–14]. Regaining hand ated in the acute stroke phase and each ESS session is dexterity (i.e., motor skills such as reaching, grasping, immediately followed by motor training of the paretic gripping, moving and releasing objects) is often achieved upper limb. Specifically, we wish to address the already within the first 4 weeks, implying that there may following: be a critical time window for recovery of upper limb functioning [9, 10] during which rehabilitation efforts (1)Does ESS treatment: (a) reduce motor and sensory may maximize functional recovery. However, there are impairments, (b) improve hand dexterity and (c) few studies investigating the effect of motor rehabilita- reduce disability at the end of the intervention tion methods in the initial weeks after stroke. period (short-term effect)? Electrical stimulation (ES) is one of the methods that (2)Are the changes that can be observed at the end of have been used to facilitate recovery of upper limb func- the intervention period still present or improved at 6 tioning following stroke. ES can induce a muscle con- months post stroke (long-term effect)? traction, or it can be a somatosensory stimulation below the motor threshold [15]. The majority of studies using Our hypothesis is that ESS treatment initiated in ES have been conducted in chronic stroke and, there- the acute stroke phase will improve paretic upper fore, it remains unknown to what extent ES applied in limb functioning as measured by the Box and Block the acute phase after stroke could affect the recovery of Test (BBT) (primary outcome measure) at 6 months upper limb functioning. Also, these investigations have post stroke. largely focused on ES that induces muscle contraction. In healthy persons, the application of low-intensity ES Methods/design with no or small motor responses to peripheral hand Trial design nerves [16–20], forearm muscles [21] or the whole hand This study is conducted as a single-blinded randomized [22, 23] elicits an increase in the cortical excitability of controlled trial with two arms and blinded endpoint ad- the representations that control the stimulated body judication. The intervention consists of ESS/sham ESS parts, which seems to outlast the stimulation period it- treatment immediately followed by training of the af- self [18, 21, 23]. It has been hypothesized that increasing fected upper limb in addition to usual rehabilitation. the amount of somatosensory input may enhance the The ESS/sham ESS treatment is initiated within the first motor recovery of patients following stroke [24]. Recent 7 days post stroke. The first 4 weeks post stroke seem to data on acute, subacute and mostly chronic stroke pa- be crucial for gaining maximal recovery of upper limb tients suggest that a single 2-h session of ESS to the per- functioning [9, 10]. Since it is not possible for us to con- ipheral hand nerves leads to transient improvement of tinue the intervention after hospital discharge for finan- pinch force, movement kinematics and upper limb cial and logistical reasons, we decided to investigate the motor skills required for ADL performance [25–31]. effect of our intervention during hospital stay (mean However, ESS was only used in conjunction with motor hospital stay when designing this study: 21 days), but training in one of these studies [29]. Interestingly, there no longer than 4 weeks post stroke. Outcome mea- is some evidence that multiple sessions of ESS to the sures are assessed at three time points: (1) within the peripheral hand nerves, in conjunction with motor first 7 days post stroke prior to intervention onset Ghaziani et al. Trials (2017) 18:84 Page 3 of 9

(baseline), (2) at hospital discharge or 4 weeks post stroke rehabilitation unit) at Bispebjerg Hospital, which is (post-intervention) and (3) at 6 months post stroke one of the university hospitals in the Capital Region (follow-up), which is the time point where the recovery of of Denmark. The hospital currently serves a well- upper limb functioning is expected to level off [12–14]. defined urban catchment area with a population of Figure 1 shows the SPIRIT flow diagram of the trial. approximately 400,000 citizens from the Municipality of Frederiksberg and the larger part of Municipality Trial setting and participants of Copenhagen for stroke rehabilitation [36]. Except Trial participants are recruited among patients admitted for holiday periods and periods of recruitment of new to the stroke unit (consisting of an acute unit and a trial personnel, all patients consecutively admitted to

Fig. 1 SPIRIT flow diagram of the trial Ghaziani et al. Trials (2017) 18:84 Page 4 of 9

thestrokeunitareassessedforeligibilitybyEGac- We are collecting data on the following variables to cording to following inclusion criteria: characterize the sample and to assess whether the randomization is successful: (a) age, (b) gender, (c) pro- 1. Admission at the rehabilitation stroke unit fession, (d) living condition (i.e., married or living with 2. Diagnosis of acute stroke confirmed by magnetic another person, alone), (e) risk factors for stroke: smok- resonance imaging (MRI) or computed tomography ing, alcohol consumption, prestroke nutritional habits, (CT) scan prestroke physical activity level, weight, height, Body 3. Upper limb paresis as indicated by a subscore <66 on Mass Index, waist circumference and other risk factors the upper limb section of the Fugl-Meyer Assessment for stroke (e.g., previous stroke, hypertension, hyperlip- (FMA) (A–D) (see “Outcome measures” section) idemia, heart disease), (f) type of stroke (i.e., hemorrhage 4. Residence in the hospital’s catchment area or infarction), (g) subtype of ischemic stroke [38], (h) se- 5. Aged 18 years or older verity of stroke at baseline [39], (i) stroke acute treatment, (j) affected upper limb, (k) dominant hand, (l) Patients are not included if any of the following exclu- medications, (m) complications during hospitalization, sion criteria are present: (n) amount and content of upper limb training following ESS and (o) amount and content of usual rehabilitation. 1. Contraindications to ESS (e.g., pacemaker in situ, For further information, see Table 1. skin impairment) [37] 2. Inability to initiate the ESS treatment within 7 days Intervention post stroke The ESS treatment is administered at the rehabilitation 3. Presence of cognitive dysfunctions or poor stroke unit. Trained health care personnel who are not communication skills in Danish that limits the involved in usual rehabilitation or outcome assessments ability to provide informed consent initiate the ESS treatment within 7 days post stroke ac- 4. Severe prestroke disability as indicated by a modified cording to a standardized protocol. Both groups receive Rankin Scale (mRS) score = 5 (see “Outcome 1 h of daily ESS from Monday to Sunday throughout measures” section) their hospital stay, but for no longer than 4 weeks post 5. Incomplete recovery of the affected upper limb after stroke. Cortical excitability in healthy persons increases a previous stroke after 2 h of suprasensory ESS of peripheral hand nerves 6. Participation in other biomedical intervention trials at the wrist [18, 20], but it is sufficient with 30 min of within the last 3 months suprasensory ESS of the whole hand to induce increased cortical excitability that lasts for 1 h after stimulation The recruitment process takes place in several steps has ended [22, 23]. When electrical stimulation is ap- which are presented in the flow diagram in Fig. 1. plied for a longer time period, it seems that cortical

Table 1 Time schedule of enrollment, intervention, assessments and responsible trial personnel Trial procedure Time points during the course of the trial Responsible personnel Blinded to group allocation? Stroke onset – 7days ESS treatment Post- intervention Hospital 6-month post stroke discharge follow-up Assess eligibility X EG Not applicable Present trial information for X EG/ESS personnel Not applicable potential participants Collect informed consent X EG Not applicable Group allocation X Administrative No/no personnel/SPM Assess outcome X X X EG Yes Collect sociodemographic data X X X EG Yes Collect medical data X X X MS/EG/ESS personnel Yes/yes/no Collect data on prestroke X EG/ESS personnel Yes/no nutritional habits and physical activity Collect data on upper limb X OTs or PTs at the stroke Yes training and usual rehabilitation unit and in the community Collect data on ESS treatment, X ESS personnel No including adverse events EG Emma Ghaziani, ESS electrical somatosensory stimulation, MS Mette Søndergaard, OT occupational therapist, PT physiotherapist, SPM S. Peter Magnusson Ghaziani et al. Trials (2017) 18:84 Page 5 of 9

excitability reaches a plateau after 45 min [19] and, any discomfort, pain or visible muscle twitches. Sham ESS therefore, we decided to provide the ESS treatment for 1 is suprasensory ESS delivered in an intermittent mode h. A Cefar Compex Theta 500 electrical device specially (active stimulation intervals of 3 s are delivered in loops of programmed to deliver ESS/sham ESS is used (DJO 2.5 min, pulse width = 250 μs, frequency = 10 Hz). In this Global Switzerland Sàrl, Ch. du Dévent, Z.I. Larges way the control group receives a total dose corresponding Pièces A, 1024 Ecublens, Switzerland). Two sets of large to 2% of the amount of active ESS delivered to the inter- electrodes (Performance electrodes, 10 × 5 cm, One vention group per ESS session, and we consider this dose Snap, DJO Nordic, AB., Murmansgatan 126, 212 25, to be too small to induce treatment effect. Apart from Malmö, Sweden) are placed on the affected arm as fol- ESS/sham ESS during the hospital stay, no other electrical lows (see Fig. 2): therapy is permitted to the affected upper limb during the 6-month trial period. Delivering other types of interven- 1. One set of electrodes on the upper arm – one tions (e.g., cognitive or motor training) is not allowed dur- electrode on the front of the shoulder and the other ing the ESS sessions. For further details, see the ESS on the back of the shoulder, both of them covering treatment protocol in Additional file 1. the lower part of the deltoid muscle The ESS/sham ESS treatment is followed by motor 2. One set of electrodes on the forearm – one electrode training of the affected upper limb that takes place just distal to the elbow crease, centered on the within 30 min after cessation of the ESS. We expect that anterior aspect of the forearm; the other electrode just brain excitability will be increased during this time inter- proximal to the palm of the hand on the anterior and val due to the ESS [18, 23]. To our knowledge, there is radial aspect of the forearm no commonly acknowledged protocol for motor training of upper limb paresis following stroke and, therefore, we The stimulation level is determined individually for decided that the motor training would be provided in each patient in the beginning of each ESS session. Half- accordance to the existing clinical practice at the depart- way into each ESS session, the stimulation level is ad- ment which is currently not standardized. Thus, the justed, if necessary, in the intervention group, whereas upper limb training consists of active, repetitive, task- the participants in the control group receive a short visit oriented practice. If the trial participant presents a se- from the ESS personnel in order to ensure that the same vere upper limb paresis with no or highly limited active amount of attention is given to both groups. movements, the therapist in charge for the particular The intervention group receives suprasensory ESS training session decides which other intervention delivered in a continuous mode (pulse width = 250 μs, methods will be employed. A task-oriented exercise bank frequency = 10 Hz). Suprasensory ESS is defined as the (see Additional file 2) is available for the physical (PTs) highest current amplitude that elicits paresthesia without and occupational therapists (OTs) delivering the upper limb training. However, it is a requirement that trial participants receive a minimum of 15 min of arm training following each ESS session. Both groups are offered the same upper limb training. Usual rehabilitation is defined as PT and OT training received by the stroke patients in the hospital as well as at rehabilitation centers and, if relevant, at nursing homes in the community during their 6-month trial participation.

Outcome measures The primary outcome measure is the Box and Block test (BBT) at 6 months post stroke. The BBT assesses the upper limb’s capacity to perform reaching, grasping, moving and releasing objects unilaterally [40]. These motor skills are essential components of performing ADLs and, therefore, of relevance when evaluating whether our intervention contributes to achieving the ultimate goal of rehabilitation: independence in daily life. Normative data are available for the healthy adult population [40], and the instrument has been validated for Fig. 2 Placement of the electrodes use in stroke patients [41]. Ghaziani et al. Trials (2017) 18:84 Page 6 of 9

The secondary outcomes measures are: (1) the upper performed after baseline assessments by contacting the extremity section of the Fugl-Meyer Assessment (FMA- administrative personnel who forwards the group alloca- UE) [42–44], (2) hand-grip strength [45], (3) palmar, tion to the personnel responsible for ESS/sham ESS lateral and thumb-to-index pinch strength [45], (4) per- treatment. SPM is occasionally involved in performance ceptual threshold of touch [46, 47], (5) pain in the upper of group allocation in the absence of the administrative limb during performance of the BBT using Numerical personnel. Rating Scale-11 [48] and (6) the mRS score [49, 50]. The FMA-UE, hand-grip strength, pinch strength, perceptual Blinding threshold of touch, and degree of pain are ways to quan- Although complete blinding of the participants to the tify the level of motor and sensory impairments in the group allocation is impossible because of the nature of upper limb. The mRS is a clinician-reported measure of the ESS treatment (i.e., participants can feel the stimula- global disability which is widely used as an endpoint in tion and are aware of whether it is delivered in continu- stroke trials. ous or intermittent sham mode), participants are kept All outcome measurements are carried out by EG. unaware of how it is supposed to feel. The personnel Baseline assessments are performed at the stroke unit. who apply ESS are not blinded to the group allocation, Post-intervention and 6-month follow-up assessments and they also collect data from medical records that will are performed at the stroke unit, the patient’s home, or be used to characterize the sample. The therapists pro- the inpatient rehabilitation center or nursing home, de- viding usual rehabilitation as well as other personnel in- pending on the residence of the participant at the sched- volved in usual patient care are blinded to the group uled time. allocation. Investigators who perform outcome assessments and data analysis are unaware of the group alloca- Sample size estimation tion, with the exception of the principal investigator Using a pretrial power analysis, we determined that a (SPM) who is involved in group allocation in the absence minimum sample size of 37 patients was required for of the administrative personnel, who are normally re- correctly detecting a within-group improvement of 5.5 sponsible for the group allocation and not involved in on the BBT, if such a difference truly exists, with a two- any other trial procedures. For further details see sided significance level of 5% and a power of 80% [51]. Table 1. The number of 5.5 blocks had previously been reported as the smallest real difference between two measure- Statistical analysis ments for the affected upper limb [41]. Since the minim- Background characteristics will be compared between ally clinically important difference on BBT is, to our the intervention groups with t tests (continuous vari- knowledge, not established yet, we made the assumption ables) or chi-squared tests (categorical variables). The that the smallest real difference would be perceived by development of both primary and secondary outcome stroke patients as being of clinical relevance for their variables will be analyzed in longitudinal models over the daily life. After adjusting for: (a) a case-fatality rate of 8% stroke recovery trajectory (baseline, post-intervention and as reported at 1 year post stroke in a Danish nationwide 6-month follow-up), and the difference of the outcomes population-based study [52] and (b) a 20% loss of partic- between the two intervention groups at each of the study ipants at follow-up for other reasons, we estimated that time points will be analyzed in multivariable linear re- 51 participants were needed in each group [53]. Hence, gression models. Analyses will be adjusted for the the total sample size is 102 participants. stratification variables (sex and the ability to perform finger extension). Possible differential attrition is ad- Randomization justed for by weighting the outcomes that are avail- Participants are allocated sequentially to either the inter- able at each of the study time points with the inverse vention or the control group using a randomization list of the probability of being present; these probabilities constructed by block randomization with variable block are estimated in logistic regression models including size. Randomization is furthermore stratified on: (a) sex all background characteristics and outcomes at previ- and (b) the ability to perform active finger extension at ous study time points. To account for this weighting baseline; active finger extension has shown to be a sim- and for repeated observations on the same individual ple and reliable early predictor of recovery of upper limb generalized estimating equations (GEE) methods are functioning in stroke patients [9, 54]. The randomization used to adjust the variance of the parameter estimates. sequence was generated with the random generator in Analyses are performed with SAS version 9.4. The statis- SAS. The randomization list is kept by administrative tical significance level is 5%. personnel and concealed from the other project investi- Additional file 3 shows the SPIRIT Checklist [55, 56] gators, with the exception of SPM. Allocations are for this study protocol. Ghaziani et al. Trials (2017) 18:84 Page 7 of 9

Discussion of the effect of ESS treatment. It is noteworthy that both To our knowledge, this is the first trial investigating the the control and the intervention group receive the same effect of multiple sessions of ESS treatment in the acute amount of attention from the personnel during the ESS stroke phase on the recovery of upper limb paresis. Sev- sessions. Although some participants might figure out eral studies have shown that the process of upper limb their group allocation, we hope that they would be moti- recovery, and especially that of hand dexterity, is most vated to continue in the trial due to fact that they are pronounced during the first 4 weeks post stroke after offered upper limb training in addition to the usual which the recovery gradually levels off before reaching a rehabilitation. plateau around 6 months post stroke. Therefore, initiation of rehabilitation in the early weeks post stroke may Trial status be essential for achieving successful upper limb recovery The recruitment of participants was initiated on 13 at the end of the rehabilitation process. Our trial intends October 2014 and will continue until complete sample to evaluate the effect of ESS – a rehabilitation method size is achieved which is expected in March 2017. At the with the potential of applicability in clinical practice. submission time of this protocol article, patients are still ESS is easy to administer, inexpensive, free of patient being recruited for the trial. discomfort and probably highly acceptable to frail patients in the early days post stroke. Additional files Contrary to previous studies that primarily focused on the stimulation of the peripheral hand nerves at Additional file 1: ESS treatment protocol version 2016_09_05. wrist level, we use two sets of large electrodes (see (DOCX 26657 kb) “Intervention” section) to stimulate somatosensory recep- Additional file 2: Exercise bank for active, repetitive, task-oriented upper limb training following stroke version 2017_01_22. (PDF 20460 kb) tors in the shoulder, elbow and wrist regions. We believe Additional file 3: The SPIRIT Checklist for the study protocol. (DOC 121 kb) that the larger area covered may be beneficial, but because we do not have a third trial arm receiving ESS at wrist Abbreviations level only, we will not be able to identify the effect of in- ADLs: Activities of daily living; ES: Electrical stimulation; ESS: Electrical creasing the stimulation areas per se. somatosensory stimulation; mRS: Modified Rankin Scale; OT: Occupational The eligibility criteria for participation in this trial are therapy/occupational therapist; PT: Physiotherapy/physiotherapist very broad. Basically, we include all adult stroke patients Acknowledgements living in the hospital’s catchment area, except those with We would like to thank: remaining upper limb paresis from a previous stroke, • the Department of Physical and Occupational Therapy for supporting this contraindications to ESS, severe prestroke disability or trial financially and nonfinancially • all PTs and OTs at the rehabilitation stroke unit for providing upper limb inability to provide informed consent. Based on this training following ESS/sham ESS and recording the training delivered to the nonrestrictive participant selection, the results can be trial participants during their hospitalization • generalized to the larger population of patients with a the nursing staff at the rehabilitation stroke unit for supporting the implementation of study procedures (e.g., ESS/sham ESS treatment, first-ever stroke or successful recovery after a previous outcome assessments) in daily clinical practice stroke and who present an upper limb paresis of varying • all the participating rehabilitation centers (Hjerneskade- og degrees of severity. However, since our trial is not pow- Rehabiliteringscentret, Rehabiliteringscenter Emdrup, Træningscenter Bispebjerg, Aktivitetscentret Bispebjerg, Træningscenter Indre By/Østerbro, ered to perform subgroup analyses, we may encounter Rehabiliteringscenter Indre By/Østerbro, Sundhedscentret Stockflethsvej, challenges in detecting the effect of ESS in such a het- Frederiksberg Rehabilitering Lioba, Frederiksbergs Kommunes erogeneous sample. The possibility of overlooking a Rehabiliteringsenhed Valby, Center for Neurorehabilitering Kurhus, Center for Sundhed og Omsorg/Sundhed og Træning Korsør, Sundhed og treatment effect on a specific subgroup of stroke patients Genoptræning Amager and the Center for Rehabilitation of Brain Injury) and (type II error) is a limitation of the study. nursing homes (Ingeborggården, Søndervang, Poppelbo) in the community Designing the control ESS treatment was challenging for recording the PT and OT training provided to our participants during the first 6 months post stroke. A special acknowledgement goes to our contact since the intervention can be perceived by the trial par- persons at each of the abovementioned institutions ticipants who may also interact with each other during • the administrative personnel at Bispebjerg Hospital responsible for group hospitalization. Since a completely inactive ESS treat- allocation • the Department of Physical and Occupational Therapy at Rigshospitalet/ ment gave rise to concerns about high dropout rates in Glostrup Hospital for introducing OTs/PTs to task-oriented upper limb training the control group, we designed a sham ESS treatment with an extremely low treatment dose. It is unlikely that Funding The trial is mainly funded by grants from the Capital Region of Denmark, the total amount of sham ESS treatment (1.2 min) is suf- Foundation for Health Research (received on 21 May 2014) and internal ficient to induce any training effects. However, we are financial, administrative and personnel support is received throughout the unable to verify this since we do not have a trial arm re- entire trial from the Department of Physical and Occupational Therapy, Bispebjerg Hospital. Other external grants from Bevica Fonden (received on ceiving completely inactive or no ESS. As a consequence, 17 December 2013), Lundbeckfonden (FP 68/2013), the Danish Association our trial results can be biased towards an underestimation of Occupational Therapists (FF 1 14–3) and Direktør Jacob Madsen and his Ghaziani et al. Trials (2017) 18:84 Page 8 of 9

wife, Olga Madsen’s Fund (5507) also contributed to the funding of this trial. 3. Jennum P, Iversen HK, Ibsen R, Kjellberg J. Cost of stroke: a controlled The funding sources have neither influence on the trial design, data collection, national study evaluating societal effects on patients and their partners. management, analysis, interpretation, nor on reporting of the results. BMC Health Serv Res. 2015;15:466. 4. Feigin VL, Barker-Collo S, McNaughton H, Brown P, Kerse N. Long-term Availability of data and materials neuropsychological and functional outcomes in stroke survivors: current – After the trial is completed and the results are disseminated in peer- evidence and perspectives for new research. Int J Stroke. 2008;3(1):33 40. reviewed journals, the datasets generated and/or analyzed during the 5. Nakayama H, Jorgensen HS, Raaschou HO, Olsen TS. Recovery of upper present trial will be deposited and publicly available at the Danish National extremity function in stroke patients: the Copenhagen Stroke Study. Arch Archives. While the trial is still ongoing, the datasets are available from the Phys Med Rehabil. 1994;75(4):394–8. corresponding author on reasonable request. 6. Lawrence ES, Coshall C, Dundas R, Stewart J, Rudd AG, Howard R, et al. Estimates of the prevalence of acute stroke impairments and disability in a multiethnic population. Stroke. 2001;32(6):1279–84. Authors’ contributions 7. Persson HC, Parziali M, Danielsson A, Sunnerhagen KS. Outcome and upper EG initiated and designed the trial, applied for funding, established extremity function within 72 hours after first occasion of stroke in an collaborative relationships with participating centers in the community, is unselected population at a stroke unit. A part of the SALGOT study. BMC responsible for trial logistics, recruitment of participants, data collection and Neurol. 2012;12:162. entry, outcome assessments, and has prepared the first version of this manuscript. CC developed the concept, designed the trial, assisted with 8. Sveen U, Bautz-Holter E, Sodring KM, Wyller TB, Laake K. Association establishing relationships with participating centers in the community, between impairments, self-care ability and social activities 1 year after – developed the ESS treatment protocol, and has critically revised this stroke. Disabil Rehabil. 1999;21(8):372 7. manuscript. CH described the ESS treatment protocol and test procedures 9. Nijland RH, van Wegen EE. Harmeling-van der Wel BC, Kwakkel G. Presence relating to the treatment, was responsible for the collaboration with the of finger extension and shoulder abduction within 72 hours after stroke clinical hospital staff, for coordinating and executing the ESS treatment predicts functional recovery: early prediction of functional outcome after – during the first year, and for the training and supervision of the additional stroke: the EPOS cohort study. Stroke. 2010;41(4):745 50. ESS personnel. VS contributed to trial design and statistical assistance, and 10. Kwakkel G, Kollen BJ, van der Grond J, Prevo AJ. Probability of regaining has critically revised this manuscript. MS contributed to trial conduct within dexterity in the flaccid upper limb: impact of severity of paresis and time – the Department of Neurology and to the collection of medical data. HC since onset in acute stroke. Stroke. 2003;34(9):2181 6. contributed to trial design and conduct within the Department of Neurology 11. Fleming MK, Newham DJ, Roberts-Lewis SF, Sorinola IO. Self-perceived and has critically revised this manuscript. SPM developed the concept, utilization of the paretic arm in chronic stroke requires high upper limb – designed the trial and edited the manuscript. All authors have read and functional ability. Arch Phys Med Rehabil. 2014;95(5):918 24. approved the final version of the manuscript. 12. Wade DT, Langton-Hewer R, Wood VA, Skilbeck CE, Ismail HM. The hemiplegic arm after stroke: measurement and recovery. J Neurol Neurosurg Psychiatry. 1983;46(6):521–4. Competing interests 13. Verheyden G, Nieuwboer A, De Wit L, Thijs V, Dobbelaere J, Devos H, et al. The authors declare that they have no competing interests. Time course of trunk, arm, leg, and functional recovery after ischemic stroke. Neurorehabil Neural Repair. 2008;22(2):173–9. Consent for publication 14. Au-Yeung SS, Hui-Chan CW. Predicting recovery of dextrous hand function Written informed consent was obtained for publication of the accompanying in acute stroke. Disabil Rehabil. 2009;31(5):394–401. images in this manuscript. The Consent Form is held by the corresponding 15. Popovic DB, Sinkaer T, Popovic MB. Electrical stimulation as a means for author and is available for review by the Editor-in-Chief. achieving recovery of function in stroke patients. NeuroRehabilitation. 2009; 25(1):45–58. Ethics approval and consent to participate 16. Ridding MC, Brouwer B, Miles TS, Pitcher JB, Thompson PD. Changes in The trial was approved by the Capital Region of Denmark’s Committee on muscle responses to stimulation of the motor cortex induced by peripheral Health Research Ethics (H-4-2014-012) on 2 October 2014 and by the Danish nerve stimulation in human subjects. Exp Brain Res. 2000;131(1):135–43. Data Protection Agency (2012-58-0004) on 21 May 2014. The trial protocol 17. Ridding MC, McKay DR, Thompson PD, Miles TS. Changes in corticomotor was notified at ClinicalTrials.gov (NCT02250365) on 18 September 2014. representations induced by prolonged peripheral nerve stimulation in Informed consent was collected from each participant before the initiation humans. Clin Neurophysiol. 2001;112(8):1461–9. of any trial procedures. 18. Kaelin-Lang A, Luft AR, Sawaki L, Burstein AH, Sohn YH, Cohen LG. Modulation of human corticomotor excitability by somatosensory input. Author details J Physiol. 2002;540(Pt 2):623–33. 1Department of Physical and Occupational Therapy, Bispebjerg Hospital, 19. McKay D, Brooker R, Giacomin P, Ridding M, Miles T. Time course of Bispebjerg Bakke 23, bldg. 10, 2400, Copenhagen, Denmark. 2Faculty of induction of increased human motor cortex excitability by nerve Health and Medical Sciences, University of Copenhagen, Copenhagen, stimulation. Neuroreport. 2002;13(10):1271–3. Denmark. 3Institute of Sports Medicine, Department of Orthopaedic Surgery 20. Wu CW, van Gelderen P, Hanakawa T, Yaseen Z, Cohen LG. Enduring M, Bispebjerg Hospital, Copenhagen, Denmark. 4Center for Healthy Aging, representational plasticity after somatosensory stimulation. Neuroimage. Faculty of Health and Medical Sciences, University of Copenhagen, 2005;27(4):872–84. 5 Copenhagen, Denmark. The Research Unit for General Practice and Section 21. Tinazzi M, Zarattini S, Valeriani M, Romito S, Farina S, Moretto G, et al. Long- of General Practice, Department of Public Health, University of Copenhagen, lasting modulation of human motor cortex following prolonged 6 Copenhagen, Denmark. Department of Neurology, Bispebjerg Hospital, transcutaneous electrical nerve stimulation (TENS) of forearm muscles: evidence Copenhagen, Denmark. of reciprocal inhibition and facilitation. Exp Brain Res. 2005;161(4):457–64. 22. Golaszewski SM, Siedentopf CM, Koppelstaetter F, Rhomberg P, Guendisch GM, Received: 4 November 2016 Accepted: 24 January 2017 Schlager A, et al. Modulatory effects on human sensorimotor cortex by whole- hand afferent electrical stimulation. Neurology. 2004;62(12):2262–9. 23. Golaszewski SM, Bergmann J, Christova M, Nardone R, Kronbichler M, References Rafolt D, et al. Increased motor cortical excitability after whole-hand 1. Murray CJ, Barber RM, Foreman KJ, Abbasoglu Ozgoren A, Abd-Allah F, electrical stimulation: a TMS study. Clin Neurophysiol. 2010;121(2):248–54. Abera SF, et al. Global, regional, and national disability-adjusted life years 24. Dobkin BH. Do electrically stimulated sensory inputs and movements (DALYs) for 306 diseases and injuries and healthy life expectancy (HALE) for lead to long-term plasticity and rehabilitation gains? Curr Opin Neurol. 188 countries, 1990–2013: quantifying the epidemiological transition. 2003;16(6):685–91. Lancet. 2015;386(10009):2145–91. 25. Koesler IB, Dafotakis M, Ameli M, Fink GR, Nowak DA. Electrical somatosensory 2. Camak DJ. Addressing the burden of stroke caregivers: a literature review. stimulation improves movement kinematics of the affected hand following J Clin Nurs. 2015;24(17–18):2376–82. stroke. J Neurol Neurosurg Psychiatry. 2009;80(6):614–9. Ghaziani et al. Trials (2017) 18:84 Page 9 of 9

26. Conforto AB, Kaelin-Lang A, Cohen LG. Increase in hand muscle strength of 50. BrunoA,ShahN,LinC,CloseB,HessDC,DavisK,etal.Improving stroke patients after somatosensory stimulation. Ann Neurol. 2002;51(1):122–5. modified Rankin Scale assessment with a simplified questionnaire. 27. Sawaki L, Wu CW, Kaelin-Lang A, Cohen LG. Effects of somatosensory Stroke. 2010;41(5):1048–50. stimulation on use-dependent plasticity in chronic stroke. Stroke. 51. University of California, San Francisco. Power and sample size programmes. 2006;37(1):246–7. [cited 28 September 2016]. Available from: https://www.stat.ubc.ca/~rollin/ 28. Klaiput A, Kitisomprayoonkul W. Increased pinch strength in acute and stats/ssize/. Accessed 4 Nov 2016. subacute stroke patients after simultaneous median and ulnar sensory 52. Langagergaard V, Palnum KH, Mehnert F, Ingeman A, Krogh BR, Bartels P, et stimulation. Neurorehabil Neural Repair. 2009;23(4):351–6. al. Socioeconomic differences in quality of care and clinical outcome after 29. Celnik P, Hummel F, Harris-Love M, Wolk R, Cohen LG. Somatosensory stroke: a nationwide population-based study. Stroke. 2011;42(10):2896–902. stimulation enhances the effects of training functional hand tasks in 53. Kirkwood BR SJ. Essential medical statistics. 2nd ed.: Blackwell Science; 2003. patients with chronic stroke. Arch Phys Med Rehabil. 2007;88(11):1369–76. 54. Smania N, Paolucci S, Tinazzi M, Borghero A, Manganotti P, Fiaschi A, et al. 30. Wu CW, Seo HJ, Cohen LG. Influence of electric somatosensory stimulation Active finger extension: a simple movement predicting recovery of arm on paretic-hand function in chronic stroke. Arch Phys Med Rehabil. function in patients with acute stroke. Stroke. 2007;38(3):1088–90. 2006;87(3):351–7. 55. Chan AW, Tetzlaff JM, Altman DG, Laupacis A, Gotzsche PC, Krleza-Jeric K, et 31. Conforto AB, Cohen LG, dos Santos RL, Scaff M, Marie SK. Effects of al. SPIRIT 2013 statement: defining standard protocol items for clinical trials. somatosensory stimulation on motor function in chronic cortico-subcortical Ann Intern Med. 2013;158(3):200–7. strokes. J Neurol. 2007;254(3):333–9. 56. Chan AW, Tetzlaff JM, Gotzsche PC, Altman DG, Mann H, Berlin JA, et al. 32. Conforto AB, Ferreiro KN, Tomasi C, dos Santos RL, Moreira VL, Marie SK, et SPIRIT 2013 explanation and elaboration: guidance for protocols of clinical al. Effects of somatosensory stimulation on motor function after subacute trials. BMJ. 2013;346:e7586. stroke. Neurorehabil Neural Repair. 2010;24(3):263–72. 33. Ikuno K, Kawaguchi S, Kitabeppu S, Kitaura M, Tokuhisa K, Morimoto S, et al. Effects of peripheral sensory nerve stimulation plus task-oriented training on upper extremity function in patients with subacute stroke: a pilot randomized crossover trial. Clin Rehabil. 2012;26(11):999–1009. 34. Dos Santos-Fontes RL, de Ferreiro Andrade KN, Sterr A, Conforto AB. Home- based nerve stimulation to enhance effects of motor training in patients in the chronic phase after stroke: a proof-of-principle study. Neurorehabil Neural Repair. 2013;27(6):483–90. 35. Laufer Y, Elboim-Gabyzon M. Does sensory transcutaneous electrical stimulation enhance motor recovery following a stroke? A systematic review. Neurorehabil Neural Repair. 2011;25(9):799–809. 36. Bispebjerg Hospital. About Bispebjerg Hospital. [cited 29 August 2016]. Available from: https://www.bispebjerghospital.dk/english/Sider/ default.aspx. Accessed 4 Nov 2016. 37. Watson Te. Electrotherapy. Evidence-based practice. 12th ed.: Elsevier; 2008. 38. Adams Jr HP, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke. 1993;24(1):35–41. 39. Multicenter trial of hemodilution in ischemic stroke—background and study protocol. Scandinavian Stroke Study Group. Stroke. 1985;16(5):885–90 40. Mathiowetz V, Volland G, Kashman N, Weber K. Adult norms for the Box and Block Test of manual dexterity. Am J Occup Ther. 1985;39(6):386–91. 41. Chen HM, Chen CC, Hsueh IP, Huang SL, Hsieh CL. Test-retest reproducibility and smallest real difference of 5 hand function tests in patients with stroke. Neurorehabil Neural Repair. 2009;23(5):435–40. 42. Fugl-Meyer AR. Post-stroke hemiplegia assessment of physical properties. Scand J Rehabil Med Suppl. 1980;7:85–93. 43. Gladstone DJ, Danells CJ, Black SE. The Fugl-Meyer Assessment of motor recovery after stroke: a critical review of its measurement properties. Neurorehabil Neural Repair. 2002;16(3):232–40. 44. Arya KN, Verma R, Garg RK. Estimating the minimal clinically important difference of an upper extremity recovery measure in subacute stroke patients. Top Stroke Rehabil. 2011;18 Suppl 1:599–610. 45. Mathiowetz V, Weber K, Volland G, Kashman N. Reliability and validity of grip and pinch strength evaluations. J Hand Surg [Am]. 1984;9(2):222–6. 46. Eek E, Engardt M. Assessment of the perceptual threshold of touch (PTT) with high-frequency transcutaneous electric nerve stimulation Submit your next manuscript to BioMed Central (Hf/TENS) in elderly patients with stroke: a reliability study. Clin Rehabil. and we will help you at every step: 2003;17(8):825–34. 47. Eek E, Holmqvist LW, Sommerfeld DK. Adult norms of the perceptual • We accept pre-submission inquiries threshold of touch (PTT) in the hands and feet in relation to age, gender, • Our selector tool helps you to ﬁnd the most relevant journal and right and left side using transcutaneous electrical nerve stimulation. • Physiother Theory Pract. 2012;28(5):373–83. We provide round the clock customer support 48. Hjermstad MJ, Fayers PM, Haugen DF, Caraceni A, Hanks GW, Loge JH, et al. • Convenient online submission Studies comparing Numerical Rating Scales, Verbal Rating Scales, and Visual • Thorough peer review Analogue Scales for assessment of pain intensity in adults: a systematic • Inclusion in PubMed and all major indexing services literature review. J Pain Symptom Manage. 2011;41(6):1073–93. 49. Banks JL, Marotta CA. Outcomes validity and reliability of the modified • Maximum visibility for your research Rankin scale: implications for stroke clinical trials: a literature review and synthesis. Stroke. 2007;38(3):1091–6. Submit your manuscript at www.biomedcentral.com/submit Paper II

78 ESS in early rehabilitation post-stroke

Electrical somatosensory stimulation in early rehabilitation of arm paresis after stroke:

A randomized controlled trial

Emma Ghaziani1, MR; Christian Couppé 1,2, PhD; Volkert Siersma3, PhD; Mette

Søndergaard4, MD; Hanne Christensen5, DMSc; S. Peter Magnusson1,2, DMSc

AUTHORS’ AFFILIATIONS

1 Department of Physical and Occupational Therapy, Bispebjerg and Frederiksberg Hospital

& Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark

2 Institute of Sports Medicine, Department of Orthopaedic Surgery M, Bispebjerg and

Frederiksberg Hospital & Center for Healthy Aging, Faculty of Health and Medical Sciences,

University of Copenhagen, Copenhagen, Denmark

3 The Research Unit for General Practice and Section of General Practice, Department of

Public Health, University of Copenhagen, Copenhagen, Denmark

4 Department of Neurology, Bispebjerg and Frederiksberg Hospital, Copenhagen, Denmark

5 Department of Neurology, Bispebjerg and Frederiksberg Hospital & Faculty of Health and

Medical Sciences, University of Copenhagen, Copenhagen, Denmark

CORRESPONDING AUTHOR: Emma Ghaziani, Bispebjerg and Frederiksberg Hospital,

Department of Physical and Occupational Therapy, Nielsine Nielsens Vej 10, 2400-

Copenhagen NW, Denmark; email: [email protected]; phone: +45 21495572.

Word count of the abstract: 250

Word count of the main text: 4,138

Number of figures and tables: 6

ESS in early rehabilitation post-stroke

ABSTRACT

Objective: To investigate the effect of electrical somatosensory stimulation (ESS) delivered during early stroke rehabilitation on the recovery of arm functioning.

Methods: 102 patients with arm paresis were randomized to a high-dose or a low-dose ESS- group within 7 days post-stroke according to our sample-size estimation. The high-dose group received 1-hour ESS to the paretic arm daily during hospitalization immediately followed by minimum 15-minute task-oriented arm training that was considered a component of the usual rehabilitation. The low-dose group received a placebo ESS followed by identical training.

Primary outcome: Box and Block Test (BBT); secondary outcomes: Fugl-Meyer Assessment

(FMA), grip strength, pinch strength, perceptual threshold of touch, pain and modified Rankin

Scale (mRS); all recorded at baseline, post-intervention and at 6 months post-stroke.

Results: There were no differences between the high-dose and the low-dose groups for any outcome measures at any time points. Improvements ≥MCID were observed for FMA, hand grip strength and mRS in both groups.

Conclusions: Providing the present ESS-protocol prior to arm training was equally beneficial as arm training alone. These results are valid for patients with mild-to-moderate stroke and moderate arm impairments. We cannot exclude benefits in patients with other characteristics, in other time intervals post-stroke or using a different ESS-protocol.

TRIAL REGISTRATION: ClinicalTrials.gov (NCT02250365)

ESS in early rehabilitation post-stroke

KEY WORDS: Acute stroke, arm paresis, hand dexterity, electrical somatosensory stimulation, early rehabilitation, recovery of arm functioning

INTRODUCTION

Stroke remains one of the leading causes of disability worldwide1. Due to an ageing population, the global, absolute number of affected individuals is increasing despite falling incidence2.

Arm paresis is one of the most frequent impairments post-stroke. Paresis is defined as a deficit in strength and motor control (the ability to perform coordinated, accurate and goal-directed movements)3. It has been reported that 48–77% of patients have arm paresis in the acute stroke phase4-6; of these, only 12–34% achieve full functional recovery at 6 months post-stroke7, 8. Arm paresis is a major obstacle in regaining independence in activities of daily living (ADL)9, especially in bimanual activities, and an almost complete functional recovery is required to routinely involve the paretic arm in ADL-performance10. Moreover, studies show that even the ability to perform ADL with the non-affected arm is lower in adults with chronic stroke compared with adults without stroke11.

The largest improvement in arm functioning occurs in the first 4–10 weeks post- stroke12-16;this is assumed to be predominantly based on spontaneous neurologic recovery.

The recovery of hand dexterity (motor skills as reaching, grasping, moving and releasing objects) appears to be determined already in the first 4 weeks post-stroke7, implying that this is a critical period for arm recovery. However, there is still limited knowledge about the effect of therapeutic interventions in the early period post-stroke when the rate of spontaneous neurologic recovery is greatest and most rehabilitation services are offered12, 17. The arm recovery continues for at least 3–6 months post-stroke4, 14, 16; subsequently, only 5–10% of the patients demonstrate further improvements14. ESS in early rehabilitation post-stroke

The potential for electrical somatosensory stimulation (ESS) to facilitate the arm recovery has been addressed in recent studies. In healthy persons, low-intensity electrical stimulation with no or minimal motor response applied to the peripheral hand nerves18-20, forearm muscles21 and the whole hand22 increases the cortical excitability of the brain areas that control the stimulated body parts, and this effect outlasts the stimulation period18, 21, 22. A single 2-hour ESS-session applied to the peripheral hand nerves transiently improves pinch force, movement kinematics and motor skills in acute, subacute and mostly chronic stroke patients23-28. Of these studies, only one used ESS in conjunction with arm training26.

Seemingly, the use of multiple ESS-sessions followed by arm training facilitates motor skills in subacute29 and chronic patients30, and the improvements appear to be long lasting in chronic patients30, but the evidence is conflicting31, 32. To our knowledge, the effect of multiple ESS-sessions in addition to arm training and initiated in the acute stroke phase was not yet investigated in a randomized controlled trial (RCT) design33. Therefore, it was justified to test the effectiveness of multiple ESS-sessions in addition to arm training on the recovery of the paretic arm when the intervention was delivered during the first 4 weeks post- stroke. ESS has no known side effects, can easily be incorporated into clinical practice, and the electrical device is relatively inexpensive.

This trial was designed to investigate the hypothesis that the application of ESS to the affected arm immediately prior to arm training during early hospitalization after stroke is superior to arm training alone regarding the recovery of hand dexterity as measured by the

Box and Block Test (BBT, primary outcome measure) at 6 months post stroke. Secondarily, we expected that improvements in dexterity would be accompanied by reductions in global disability, and motor and sensory impairments. Thus, we planned to address the following research questions: a) Does ESS reduce motor and sensory impairments, improve hand dexterity, and reduce disability at the end of the ESS-intervention period?; b) Are possible ESS in early rehabilitation post-stroke benefits observed at the end of the intervention period still present or improved at 6 months post-stroke?

METHODS

A detailed description of the trial protocol has been published elsewhere34.

Trial design

The trial was conducted as an RCT with blinded outcome adjudication and two parallel arms: an intervention and a control group. The trial was notified at ClinicalTrials.gov

(NCT02250365).

Trial setting and participants

The trial was conducted in the stroke rehabilitation unit of Bispebjerg Hospital,

Copenhagen, Denmark. After a few days in the acute stroke unit the patients were transferred to the stroke rehabilitation unit where they received early inpatient rehabilitation while they still needed constant medical care. Subsequently, the patients were discharged to their homes with/without referral to outpatient rehabilitation services in the community, inpatient rehabilitation centers in the community, or nursing homes, depending on their recovery potential and needed ADL-assistance. All these rehabilitation services are covered by the

Danish national health insurance.

The eligibility criteria were: a) age ≥18 years; b) residence in the hospital’s catchment area for stroke rehabilitation; c) acute stroke confirmed by magnetic resonance imaging (MRI) or computer tomography (CT) scan; and d) arm paresis (<66 points on the upper extremity section of the Fugl-Meyer Assessment, subscale A–D). Patients were excluded if any of the following criteria were present: a) contraindications to ESS ESS in early rehabilitation post-stroke

(pacemaker, skin impairment); b) inability to initiate the ESS within 7 days post-stroke due to medical or logistical issues; c) cognitive dysfunctions or poor Danish communication skills that limited the ability to provide informed consent; d) severe pre-stroke disability (modified

Rankin Scale score=5; e) incomplete recovery of the affected arm after a previous stroke; and f) participation in other biomedical intervention trials within the last 3 months.

Data on following potential demographic and clinical confounders were collected from medical records: a) age; b) sex; c) living arrangement; d) profession35; e) stroke risk factors: smoking, high risk of disease due to alcohol consumption36, increased to high risk of metabolic disease due to Body Mass Index (BMI) and waist circumference37, previous stroke and other diseases, pre-stroke physical activity level (Physical Activity Scale for the Elderly38); f) type of stroke (ischemic stroke, IS; hemorrhagic stroke, HS); g) subtype of IS (TOAST classification39); h) stroke severity (Scandinavian Stroke Scale40, 41); i) voluntary finger extension8; j) affected arm; k) dominant hand; l) acute stroke treatment; m) medications at hospital admission; n) blood pressure and pulse on second day after hospital admission; o) blood test results on the hospital admission day or following day. Moreover, we gathered data on complications during hospitalization, discharge destination, residence at 6 months post-stroke, recurrent stroke, and the actual practice time and content of usual rehabilitation (including arm training).

Participants provided written informed consent before enrolment. Ethical approval was obtained from the Capital Region of Denmark’s Committee on Health Research

Ethics (H-4-2014-012) and by the Danish Data Protection Agency (2012-58-0004).

Intervention

The ESS was initiated within the first 7 days post-stroke, and administered by trained health care personnel according to the published protocol, Additional file 134. Two ESS in early rehabilitation post-stroke sets of external electrodes were positioned at the wrist, elbow and shoulder level. The electrode at the wrist level covered all three peripheral hand nerves; the stimulation at the elbow and shoulder level was of cutaneous art. The intervention group received supra-sensory

ESS delivered in continuous mode (pulse width=250µs, frequency=10Hz). Supra-sensory

ESS was defined as the highest current amplitude that elicits paresthesia without any discomfort, pain or visible muscle twitches. Because supra-sensory ESS can be perceived and the participants could communicate to each other during hospitalization, using a completely inactive placebo ESS gave rise to concerns about a possible high drop-out in the control group. Consequently, we designed a placebo ESS with a low treatment dose that we considered unlikely to induce any training effects. The placebo ESS consisted of suprasensory ESS delivered in intermittent mode (active stimulation intervals of 3s delivered in loops of 2.5 minutes, pulse width=250µs, frequency=10Hz), and the control group received a total dose corresponding to 2% of the amount of active ESS delivered to the intervention group per ESS-session (see Supplementary). The groups received 1 hour of daily high-dose

/low-dose ESS from Monday to Sunday throughout the hospital stay, but no longer than 4 weeks post-stroke. The stimulation level was determined individually for each participant at the beginning of each ESS-session. To avoid adaptation to ESS in the high-dose group, the stimulation level was adjusted, if necessary, after the first 30 minutes, whereas the low-dose group got a short visit from the ESS-staff to ensure the same amount of attention. Patients were neither requested to focus on the stimulation, nor allowed to participate in other training sessions (e.g. speech and language therapy, gait training) during ESS-sessions. The high- dose/low-dose ESS was followed by minimum 15-minute arm training; this intervention component was considered a part of the usual inpatient rehabilitation. The arm training consisted of active, repetitive, task-oriented practice and was provided within 30 minutes after cessation of ESS when we expected the brain excitability to be increased18, 20. If the ESS in early rehabilitation post-stroke participants were unable to perform volitional arm movements, the treating therapist decided which training methods to be employed. Except from an inspirational task-oriented exercise bank34, the arm training was not further standardized. The high-dose/low-dose ESS was provided in addition to a) usual rehabilitation defined as physical (PT) and occupational therapy (OT) delivered in the hospital’s rehabilitation stroke unit, rehabilitation centers in the community, nursing homes, and b) other physical stroke rehabilitation (e.g. PT services in private clinics) during their trial participation.

Outcomes

Our primary outcome measure was BBT at 6 months post-stroke. Normative data are available for the healthy adult population42 and the instrument has been validated for use in stroke patients43. To our knowledge, minimally clinically important difference (MCID) for BBT has not been established yet; the minimal detectable change (MDC) is 5.5 blocks/min43.

The secondary outcomes measures were: a) upper extremity section of the Fugl-

Meyer Assessment, subscale A−D (FMA-UE-AD) (English version)44; b) hand grip strength45; c) palmar, key and tip pinch strength45; d) perceptual threshold of touch (PTT)46; e) pain in the arm during performance of BBT using Numerical Rating Scale-1147; and f) modified Rankin Scale (mRS)48 (score 0−5). MCID have been reported for: a) FMA-UE: 10 points49; b) hand grip strength: 5.0kg (dominant hand affected) and 6.2kg (non-dominant hand affected)50, and c) mRS: ≥1 point51.

The outcome measures were assessed at three time-points by EG: a) prior to intervention onset (baseline); b) at hospital discharge or 4 weeks post-stroke (post- intervention); and c) at 6 months post-stroke (follow-up). All baseline assessments were performed at the stroke unit. Post-intervention and follow-up assessments were performed at ESS in early rehabilitation post-stroke the participants’ residences at that specific point in time (stroke unit, patient’s home, inpatient rehabilitation centers, nursing homes). All outcome measures were regarded as numerical scales.

Sample size

Our pre-trial sample-size analysis showed that 37 patients were required for detecting a within-group improvement corresponding to the reported MDC=5.5 blocks/min43 on BBT with a two-sided significance level=5% and a power=80%. After adjusting for a case- fatality rate=8% and an estimated drop-out=20% for other reasons, 51 participants were needed in each group, which resulted in a total sample size=102 participants.

Randomization

Participants were allocated sequentially to the low-dose or the high-dose group using a randomization list constructed by block randomization with variable block size.

Randomization was stratified by: a) sex and b) ability to perform voluntary finger extension8 at baseline. The randomization sequence was generated with the random generator in SAS version 9.4. Until data analysis was completed, the randomization list was kept by administrative staff and concealed from OTs/PTs responsible for arm training, assessor (EG) and data analysts (VS, EG). The administrative staff allocated the participants to one of the trial groups after baseline assessments and then forwarded the result by email to the ESS- staff. In the absence of the administrative staff, the principal investigator (SPM) was occasionally involved in assigning participants to trial groups.

Blinding ESS in early rehabilitation post-stroke

Complete blinding of participants was not feasible as the participants could perceive the stimulation and figure out their group allocation. However, the participants were kept unaware of which stimulation mode was the intervention and which one functioned as control. The ESS-staff was not blinded to the group allocation. The therapists providing usual rehabilitation, the staff involved in medical care, the outcome assessor (EG) and data analysts

(VS, EG) were kept blinded to group allocation until all analyses were completed.

Statistical methods

The demographic and clinical characteristics at baseline were compared between the groups with Fisher’s exact test for categorical variables and with Wilcoxon test for continuous variables as they did not follow a normal distribution. The development of primary and secondary outcome variables was analyzed in longitudinal models over the stroke recovery trajectory (baseline, post-intervention and follow-up), and the differences of the outcomes between the two groups at each of the trial time points were analysed in multivariable linear regression models. Analyses were adjusted for the stratification variables

(sex and the ability to perform voluntary finger extension at baseline). An informal inspection of the summary statistics of the missing participants showed some general trends: they were mostly women, had the dominant hand affected and a pre-stroke mRS>0. To adjust for possible differential dropout, the available data at each of the trial time points was weighted with the inverse probability of being observed; these probabilities were estimated from logistic regression models with sex, affected dominant hand, pre-stroke mRS and outcomes at previous trial points as covariates. To account for this weighting and for repeated observations on the same participant, generalized estimating equations (GEE) methods are used to adjust the variance of the parameter estimates. For the hand grip strength and pinch strength, the maximum value of the 3–5 measurements done per assessment was entered in the data set. ESS in early rehabilitation post-stroke

For the PTT, 3–5 measurements per assessment were undertaken, and the value entered in the data set was the mean of the remaining measurements after the lowest and the highest values were discarded. BBT was performed once per assessment. Because baseline data on mRS were not collected, we assigned a baseline mRS of (4+5)/2=4.5 points to all the trial participants. This estimation is supported by the assumption that hospitalized stroke patients have a mRS of 4 or 5 due to their need for constant medical care. Analyses were performed with SAS version 9.4. The statistical significance level was set to 1% to guard against false detection of effects because of multiple comparisons. For differences in our primary outcome, the significance level was of 5%.

RESULTS

Figure 1 details the participants’ flow through the trial. All patients consecutively admitted to the stroke rehabilitation unit and diagnosed with IS or HS from

October 2014 to March 2017 were screened for a newly developed arm paresis, except a total of 6 months (holidays, recruitment/training of new trial staff) (n=1,214). More than 537 stroke patients were identified (the screening log for the initial 5 months was incomplete). Of these,

102 patients were randomized; 49 to the low-dose group and 53 to the high-dose group. Two patients were assigned to the low-dose group although it was not possible to initiate the ESS within 7 days post-stroke; 1 patient was assigned to the high-dose group despite a pre-stroke mRS=5 and inability to initiate the ESS within 7 days post-stroke. All randomized participants received the assigned intervention. All baseline assessments were conducted during the first 7 days post-stroke. Post-intervention data were collected approximately 1 day after the last ESS-session and 19–20 days after stroke onset from 45 participants in the low- dose group and from all 53 participants in the high-dose group. Follow-up assessments were collected approximately 3.5−5 days after the actual date for 6 months post-stroke from 42 ESS in early rehabilitation post-stroke participants in the low-dose group, and 46 participants in the high-dose group (Table 1). Data from all randomized participants were entered in the intention-to-treat analysis.

Figure 1

Table 1

The two groups were comparable with respect to demographic and clinical characteristics at baseline (Table 2). Participants exhibited a mean FMA-UE-AD of 33.2−33.6 points at baseline, indicating a moderate arm impairment52.

Table 2

Our analysis showed no statistically significant between-group differences for any outcome measures at any time points (Table 3). However, there were positive, statistically significant within-group effects in both trial groups on all outcome measures, except for pain; improvements in FMA-UE and mRS were moreover ≥MCID at all time points, and for the hand grip strength at follow-up (Table 4).

Table 3

Table 4

During their hospital stay (17−18 days) (Table 5), the participants received

12−13 ESS-interventions; of these, 9−10 were given per protocol. The amount of arm training during the ESS-intervention period was of 4−5.3 hours and the total amount of arm training during the whole trial was of 10.1−15.7 hours. During the ESS-intervention period, the ESS in early rehabilitation post-stroke amount of additional OT/PT training was 5.8−7.1 hours and the total amount of additional

OT/PT training during the whole trial was 30.9−34.4 hours. Around 23% of the participants in both groups also received other stroke rehabilitation, mostly consisting of individually purchased PT services. Table 1 presents detailed information on the delivered ESS- intervention and usual rehabilitation (including arm training), showing similarity between the groups with respect to these parameters.

Adverse reactions (itch) to ESS were reported in 1 patient in each group and alleviated by using allergy friendly electrodes. The percentage of complications during the hospitalization was similar in both groups. Likewise, the percentage of recurrent stroke was comparable; 6% in the low-dose group and 4% in the high-dose group. The ESS-intervention had no effect on the discharge destination or the participants’ residence at follow-up (Table

5).

Table 5

DISCUSSION

This RCT investigated the effect of ESS on arm recovery and demonstrated that adding the present ESS-protocol to arm training in the first 4 weeks post-stroke is equally effective as arm training alone in improving hand dexterity, motor and sensory functions, and global disability. Both trial groups showed a positive, statistically significant effect of time for the most outcome measures; moreover, improvements in motor function, hand grip strength and global disability were ≥MCID.

Several studies have demonstrated a positive, albeit transient, effect of a single

2-hour ESS-session on different aspects of upper limb functioning in all stroke phases23-28. A systematic review with meta-analyses 53 including trials with repetitive ESS-sessions could ESS in early rehabilitation post-stroke not demonstrate any convincing, beneficial effect in chronic stroke. Importantly, this conclusion rests on a very limited amount of data. Recently, an RCT54 in subacute/chronic stroke patients with some active finger and wrist extension demonstrated positive, long-term effects in favour of ESS on functional capacity, but not on motor functions and the Stroke

Impact Scale (SIS). Interestingly, the SIS was positively influenced by the inactive placebo

ESS, but the effect was temporary. This trial stimulated the radial and ulnar nerves and applied a high intervention dose (18 sessions x 2-hour ESS followed by 4-hour task-oriented arm training). To our knowledge, only one other trial55 except the present one has been conducted in the acute stroke phase. Contrary to our results, it showed a positive long-term effect of ESS on hand grip strength and tip pinch strength in early rehabilitation of patients with mild-to-moderate stroke and severe arm impairments. It should be noted that the intervention dose was higher (20 sessions x 60-minute ESS) and the stimulation protocol was different compared with ours (200µs pulses at 20Hz with maximum tolerated intensity to acupuncture points); ESS was given in addition to usual rehabilitation, but not specifically followed by arm training. In subacute stroke, one study29 reported beneficial short-term effect of 12 sessions x 20-minute task-oriented practice preceded by 2-hour subsensory ESS to the median nerve. Furthermore, a post-intervention positive effect was also shown after 10 sessions x 45-minute ESS delivered by a stimulation glove supplementary to usual rehabilitation56. Therefore, it appears that the current evidence for ESS during the acute stroke phase is conflicting. It also remains unknown if ESS may play a role in later stroke phases, and if so under which conditions. Interestingly, a recent systematic review with meta- analyses57 demonstrated an advantageous effect of ESS when applied on to the lower limb, especially in the acute/subacute stroke.

In the current trial, 1-hour ESS was provided in conjunction with minimum 15- minute task-oriented arm training, which has previously shown to be an effective intervention ESS in early rehabilitation post-stroke during the first 4 weeks post-stroke58. When used early after stroke, it is possible that ESS is not sufficiently potent to induce improvements beyond spontaneous neurologic recovery and the effect of task-oriented arm training. Note that the above-mentioned trial in acute stroke55 did not combined ESS with task-oriented arm training. A second possible explanation for the lack of effect in our trial is the relatively low number of ESS-sessions (12−13) compared with higher number (18−28) in other studies30, 54, 55. Unfortunately, financial and logistical reasons prevented an ESS-intervention beyond 4 weeks post-stroke in this trial, and we cannot rule out that a longer intervention period might have altered the results. Likewise, we cannot rule out that using a standardized arm training protocol in which interventions such as neuromuscular electrical stimulation or robot-assisted therapy59 prescribed to patients with severe impairments could have influenced the outcomes. Thirdly, we cannot exclude the possibility of a placebo effect, which has also been suggested by other55, or of our placebo intervention having a real effect. The influence of electrode location and stimulation parameters is difficult to assess as prior stimulation protocols vary considerably33, 53.

Although we demonstrated within-group improvements ≥MCID, it is important to emphasize that it is unknown whether they translated into an increased use of the affected arm in real-life situations. We would recommend that future studies employ, for example, accelerometers to capture the impact of arm interventions on real-life ADL-performance, as well as participation measures that provide information on whether the intervention affects the patient’s fulfilment of life roles.

Stroke recovery can be defined as improved performance without distinguishing between the degree of compensation and pure neurologic recovery caused by physiological changes. In this trial, we applied outcome measures that capture both aspects of recovery: pinch strength, hand grip strength, PTT and FMA-UE for pure neurologic recovery, and BBT and mRS for recovery of functional capacity that may partially be obtained through ESS in early rehabilitation post-stroke compensatory strategies. However, we acknowledge that clinical measures as our selected primary outcome (BBT) cannot distinguish between pure neurologic recovery and compensation. To address this limitation, we support the recently published consensus-based recommendations for measurement in stroke recovery and rehabilitation trials 60 in urging the use of kinematic and kinetic measures in future intervention trials for the upper limb after stroke.

There was a trend that the low-dose group received a lower dose of arm training during the ESS-intervention period (4h vs. 5.3h, p=0.02) (Table 1) and comprised more participants with a severe stroke compared with the high-dose group (8.2% vs. 0%, p=0.05)

(Table 2), though without reaching the significance level of 1% applied for all the comparisons, except those regarding our primary outcome measure. However, the low-dose group improved similarly to the high-dose group. One possible explanation is the existence of a placebo or a real effect of the low-dose ESS; a second explanation is that the differences between the trial groups are too narrow to have any clinical importance. Conversely, we cannot be sure that the participants with initial severe arm paresis are equally distributed between the trial groups regarding their potential for recovery61 because we did not use transcranial magnetic stimulation (TMS) and MRI to assess the degree of structural neural damage.

Our eligibility criteria were broad. However, the final sample was homogenous, comprising mostly of patients with mild-to-moderate stroke, moderate arm impairments, and some voluntary finger extension. Consequently, the generalizability of our results is limited to stroke patients with these characteristics. Because the characteristics of the patients excluded from the trial (Figure 1) were not recorded, we cannot assess whether these patients were significantly different from those enrolled in the trial. Finally, the trial was not powered to perform subgroup analyses. Therefore, we were not able to detect possible effects on ESS in early rehabilitation post-stroke subgroups based on, for example, severity of arm paresis or potential for recovery based on

TMS and MRI61.

CONCLUSION

Using the ESS-protocol described herein immediately followed by arm training during the first 4 weeks post-stroke does not add to arm training alone in enhancing the recovery of arm functioning. These results apply for patients with mild-to-moderate stroke, moderate arm impairments and some voluntary finger extension. Whether ESS initiated early after stroke is effective in patients with severe-to-moderate stroke and/or with potential for recovery despite initial inability of finger extension remains unknown. Likewise, whether

ESS is effective in later time intervals after stroke and under which conditions (electrode placement, stimulation parameters, intervention dose, timing in relation to arm training) requires further investigation.

FUNDING

This work was supported by the Capital Region of Denmark, Foundation for

Health Research, Bevica Fonden, Lundbeck Foundation, the Danish Association of

Occupational Therapists, Direktør Jacob Madsen & hustru Olga Madsen’s fond, and the

Department of Physical and Occupational Therapy at Bispebjerg Hospital. The funding sources had no influence on the trial design, data collection, analyses, interpretation and reporting of results.

CONFLICTING INTERESTS

The authors declare no conflict of interest.

ESS in early rehabilitation post-stroke

AVAILABILITY OF DATA

The data set is available from the corresponding author on reasonable request.

REFERENCES

1. Feigin VL, Krishnamurthi RV, Parmar P, et al. Update on the Global Burden of Ischemic and Hemorrhagic Stroke in 1990-2013: The GBD 2013 Study. Neuroepidemiology 2015; 45: 161-176. 2015/10/28. DOI: 10.1159/000441085. 2. Roth GA, Johnson C, Abajobir A, et al. Global, Regional, and National Burden of Cardiovascular Diseases for 10 Causes, 1990 to 2015. J Am Coll Cardiol 2017; 70: 1-25. 2017/05/22. DOI: 10.1016/j.jacc.2017.04.052. 3. Cortes JC, Goldsmith J, Harran MD, et al. A Short and Distinct Time Window for Recovery of Arm Motor Control Early After Stroke Revealed With a Global Measure of Trajectory Kinematics. Neurorehabilitation and neural repair 2017; 31: 552-560. 2017/05/17. DOI: 10.1177/1545968317697034. 4. Nakayama H, Jorgensen HS, Raaschou HO, et al. Recovery of upper extremity function in stroke patients: the Copenhagen Stroke Study. Arch Phys Med Rehabil 1994; 75: 394-398. 1994/04/01. 5. Lawrence ES, Coshall C, Dundas R, et al. Estimates of the prevalence of acute stroke impairments and disability in a multiethnic population. Stroke 2001; 32: 1279-1284. 2001/06/02. 6. Persson HC, Parziali M, Danielsson A, et al. Outcome and upper extremity function within 72 hours after first occasion of stroke in an unselected population at a stroke unit. A part of the SALGOT study. BMC neurology 2012; 12: 162. 2013/01/01. DOI: 10.1186/1471-2377-12-162. 7. Kwakkel G, Kollen BJ, van der Grond J, et al. Probability of regaining dexterity in the flaccid upper limb: impact of severity of paresis and time since onset in acute stroke. Stroke 2003; 34: 2181-2186. 2003/08/09. DOI: 10.1161/01.str.0000087172.16305.cd. 8. Nijland RH, van Wegen EE, Harmeling-van der Wel BC, et al. Presence of finger extension and shoulder abduction within 72 hours after stroke predicts functional recovery: early prediction of functional outcome after stroke: the EPOS cohort study. Stroke 2010; 41: 745-750. 2010/02/20. DOI: 10.1161/strokeaha.109.572065. 9. Sveen U, Bautz-Holter E, Sodring KM, et al. Association between impairments, self- care ability and social activities 1 year after stroke. Disability and rehabilitation 1999; 21: 372-377. 1999/09/30. 10. Fleming MK, Newham DJ, Roberts-Lewis SF, et al. Self-perceived utilization of the paretic arm in chronic stroke requires high upper limb functional ability. Arch Phys Med Rehabil 2014; 95: 918-924. 2014/02/01. DOI: 10.1016/j.apmr.2014.01.009. 11. Michielsen ME, Selles RW, Stam HJ, et al. Quantifying nonuse in chronic stroke patients: a study into paretic, nonparetic, and bimanual upper-limb use in daily life. Arch Phys Med Rehabil 2012; 93: 1975-1981. 2012/04/03. DOI: 10.1016/j.apmr.2012.03.016. 12. van Kordelaar J, van Wegen E and Kwakkel G. Impact of time on quality of motor control of the paretic upper limb after stroke. Arch Phys Med Rehabil 2014; 95: 338-344. 2013/10/29. DOI: 10.1016/j.apmr.2013.10.006. 13. Kwakkel G, Kollen B and Twisk J. Impact of time on improvement of outcome after stroke. Stroke 2006; 37: 2348-2353. 2006/08/26. DOI: 10.1161/01.STR.0000238594.91938.1e. 14. Kwakkel G and Kollen BJ. Predicting activities after stroke: what is clinically relevant? International journal of stroke : official journal of the International Stroke Society 2013; 8: 25-32. DOI: 10.1111/j.1747-4949.2012.00967.x. ESS in early rehabilitation post-stroke

15. Lee KB, Lim SH, Kim KH, et al. Six-month functional recovery of stroke patients: a multi-time-point study. Int J Rehabil Res 2015; 38: 173-180. 2015/01/21. DOI: 10.1097/mrr.0000000000000108. 16. Verheyden G, Nieuwboer A, De Wit L, et al. Time course of trunk, arm, leg, and functional recovery after ischemic stroke. Neurorehabilitation and neural repair 2008; 22: 173-179. 2007/09/19. DOI: 10.1177/1545968307305456. 17. Higgins J, Mayo NE, Desrosiers J, et al. Upper-limb function and recovery in the acute phase poststroke. Journal of rehabilitation research and development 2005; 42: 65-76. 2005/03/03. 18. Kaelin-Lang A, Luft AR, Sawaki L, et al. Modulation of human corticomotor excitability by somatosensory input. The Journal of physiology 2002; 540: 623-633. 2002/04/17. 19. McKay D, Brooker R, Giacomin P, et al. Time course of induction of increased human motor cortex excitability by nerve stimulation. Neuroreport 2002; 13: 1271-1273. 2002/08/02. 20. Wu CW, van Gelderen P, Hanakawa T, et al. Enduring representational plasticity after somatosensory stimulation. NeuroImage 2005; 27: 872-884. 2005/08/09. DOI: 10.1016/j.neuroimage.2005.05.055. 21. Tinazzi M, Zarattini S, Valeriani M, et al. Long-lasting modulation of human motor cortex following prolonged transcutaneous electrical nerve stimulation (TENS) of forearm muscles: evidence of reciprocal inhibition and facilitation. Experimental brain research 2005; 161: 457-464. 2004/11/20. DOI: 10.1007/s00221-004-2091-y. 22. Golaszewski SM, Siedentopf CM, Koppelstaetter F, et al. Modulatory effects on human sensorimotor cortex by whole-hand afferent electrical stimulation. Neurology 2004; 62: 2262-2269. 2004/06/24. 23. Koesler IB, Dafotakis M, Ameli M, et al. Electrical somatosensory stimulation improves movement kinematics of the affected hand following stroke. Journal of neurology, neurosurgery, and psychiatry 2009; 80: 614-619. 2008/11/18. DOI: 10.1136/jnnp.2008.161117. 24. Sawaki L, Wu CW, Kaelin-Lang A, et al. Effects of somatosensory stimulation on use- dependent plasticity in chronic stroke. Stroke 2006; 37: 246-247. 2005/12/03. DOI: 10.1161/01.STR.0000195130.16843.ac. 25. Klaiput A and Kitisomprayoonkul W. Increased pinch strength in acute and subacute stroke patients after simultaneous median and ulnar sensory stimulation. Neurorehabilitation and neural repair 2009; 23: 351-356. 2008/11/05. DOI: 10.1177/1545968308324227. 26. Celnik P, Hummel F, Harris-Love M, et al. Somatosensory stimulation enhances the effects of training functional hand tasks in patients with chronic stroke. Arch Phys Med Rehabil 2007; 88: 1369-1376. 2007/10/30. DOI: 10.1016/j.apmr.2007.08.001. 27. Wu CW, Seo HJ and Cohen LG. Influence of electric somatosensory stimulation on paretic-hand function in chronic stroke. Arch Phys Med Rehabil 2006; 87: 351-357. 2006/02/28. DOI: 10.1016/j.apmr.2005.11.019. 28. Conforto AB, Cohen LG, dos Santos RL, et al. Effects of somatosensory stimulation on motor function in chronic cortico-subcortical strokes. Journal of neurology 2007; 254: 333-339. 2007/03/09. DOI: 10.1007/s00415-006-0364-z. 29. Conforto AB, Ferreiro KN, Tomasi C, et al. Effects of somatosensory stimulation on motor function after subacute stroke. Neurorehabilitation and neural repair 2010; 24: 263-272. 2009/11/04. DOI: 10.1177/1545968309349946. 30. Dos Santos-Fontes RL, Ferreiro de Andrade KN, Sterr A, et al. Home-based nerve stimulation to enhance effects of motor training in patients in the chronic phase after stroke: a proof-of-principle study. Neurorehabilitation and neural repair 2013; 27: 483-490. 2013/03/13. DOI: 10.1177/1545968313478488. 31. Ikuno K, Kawaguchi S, Kitabeppu S, et al. Effects of peripheral sensory nerve stimulation plus task-oriented training on upper extremity function in patients with subacute stroke: a pilot randomized crossover trial. Clinical rehabilitation 2012; 26: 999-1009. 2012/04/14. DOI: 10.1177/0269215512441476. ESS in early rehabilitation post-stroke

32. Sullivan JE, Hurley D and Hedman LD. Afferent stimulation provided by glove electrode during task-specific arm exercise following stroke. Clinical rehabilitation 2012; 26: 1010-1020. Randomized Controlled Trial

Research Support, Non-U.S. Gov't 2012/05/09. DOI: 10.1177/0269215512442915. 33. Laufer Y and Elboim-Gabyzon M. Does sensory transcutaneous electrical stimulation enhance motor recovery following a stroke? A systematic review. Neurorehabilitation and neural repair 2011; 25: 799-809. 2011/07/13. DOI: 10.1177/1545968310397205. 34. Ghaziani E, Couppe C, Henkel C, et al. Electrical somatosensory stimulation followed by motor training of the paretic upper limb in acute stroke: study protocol for a randomized controlled trial. Trials 2017; 18: 84. DOI: 10.1186/s13063-017-1815-9. 35. Hedblad B, Jonsson S, Nilsson P, et al. Obesity and myocardial infarction--vulnerability related to occupational level and marital status. A 23-year follow-up of an urban male Swedish population. Journal of internal medicine 2002; 252: 542-550. 2002/12/11. 36. Danish Health Authority. Høj risiko for sygdom på grund af alkohol, https://www.sst.dk/da/sundhed-og-livsstil/alkohol/anbefalinger/hoej-risiko-for-sygdom (accessed January 3 2018). 37. WHO. Waist Circumference and Waist–Hip Ratio: Report of a WHO Expert Consultation, http://apps.who.int/iris/bitstream/10665/44583/1/9789241501491_eng.pdf (2008, accessed 17 November 2017). 38. Washburn RA, Smith KW, Jette AM, et al. The Physical Activity Scale for the Elderly (PASE): development and evaluation. J Clin Epidemiol 1993; 46: 153-162. 1993/02/01. 39. Adams HP, Jr., Bendixen BH, Kappelle LJ, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993; 24: 35-41. 1993/01/01. 40. Multicenter trial of hemodilution in ischemic stroke--background and study protocol. Scandinavian Stroke Study Group. Stroke 1985; 16: 885-890. 1985/09/01. 41. Boysen G and Christensen H. Stroke severity determines body temperature in acute stroke. Stroke 2001; 32: 413-417. 2001/02/07. 42. Mathiowetz V, Volland G, Kashman N, et al. Adult norms for the Box and Block Test of manual dexterity. The American journal of occupational therapy : official publication of the American Occupational Therapy Association 1985; 39: 386-391. 1985/06/01. 43. Chen HM, Chen CC, Hsueh IP, et al. Test-retest reproducibility and smallest real difference of 5 hand function tests in patients with stroke. Neurorehabilitation and neural repair 2009; 23: 435-440. 2009/03/06. DOI: 10.1177/1545968308331146. 44. Fugl-Meyer AR. Post-stroke hemiplegia assessment of physical properties. Scandinavian journal of rehabilitation medicine Supplement 1980; 7: 85-93. 1980/01/01. 45. Mathiowetz V, Weber K, Volland G, et al. Reliability and validity of grip and pinch strength evaluations. The Journal of hand surgery 1984; 9: 222-226. 1984/03/01. 46. Eek E and Engardt M. Assessment of the perceptual threshold of touch (PTT) with high-frequency transcutaneous electric nerve stimulation (Hf/TENS) in elderly patients with stroke: a reliability study. Clinical rehabilitation 2003; 17: 825-834. 2003/12/20. 47. Hjermstad MJ, Fayers PM, Haugen DF, et al. Studies comparing Numerical Rating Scales, Verbal Rating Scales, and Visual Analogue Scales for assessment of pain intensity in adults: a systematic literature review. J Pain Symptom Manage 2011; 41: 1073-1093. 2011/05/31. DOI: 10.1016/j.jpainsymman.2010.08.016. 48. Banks JL and Marotta CA. Outcomes validity and reliability of the modified Rankin scale: implications for stroke clinical trials: a literature review and synthesis. Stroke; a journal of cerebral circulation 2007; 38: 1091-1096. 2007/02/03. DOI: 10.1161/01.STR.0000258355.23810.c6. 49. Arya KN, Verma R and Garg RK. Estimating the minimal clinically important difference of an upper extremity recovery measure in subacute stroke patients. Topics in stroke rehabilitation 2011; 18 Suppl 1: 599-610. 2011/11/29. DOI: 10.1310/tsr18s01-599. ESS in early rehabilitation post-stroke

50. Lang CE, Edwards DF, Birkenmeier RL, et al. Estimating minimal clinically important differences of upper-extremity measures early after stroke. Arch Phys Med Rehabil 2008; 89: 1693- 1700. 2008/09/02. DOI: 10.1016/j.apmr.2008.02.022. 51. Dromerick AW, Edwards DF and Diringer MN. Sensitivity to changes in disability after stroke: a comparison of four scales useful in clinical trials. Journal of rehabilitation research and development 2003; 40: 1-8. 2004/05/20. 52. Pang MY, Harris JE and Eng JJ. A community-based upper-extremity group exercise program improves motor function and performance of functional activities in chronic stroke: a randomized controlled trial. Arch Phys Med Rehabil 2006; 87: 1-9. 2006/01/13. DOI: 10.1016/j.apmr.2005.08.113. 53. Grant VM, Gibson A and Shields N. Somatosensory stimulation to improve hand and upper limb function after stroke-a systematic review with meta-analyses. Topics in stroke rehabilitation 2017: 1-11. 2017/10/21. DOI: 10.1080/10749357.2017.1389054. 54. Carrico C, Westgate PM, Powell ES, et al. Nerve Stimulation Enhances Task-Oriented Training for Moderate-to-Severe Hemiparesis 3-12 Months after Stroke: A Randomized Trial. American journal of physical medicine & rehabilitation 2018 2018/05/26. DOI: 10.1097/PHM.0000000000000971. 55. Au-Yeung SS and Hui-Chan CW. Electrical acupoint stimulation of the affected arm in acute stroke: a placebo-controlled randomized clinical trial. Clinical rehabilitation 2014; 28: 149-158. 2013/08/02. DOI: 10.1177/0269215513494875. 56. Kattenstroth JC, Kalisch T, Sczesny-Kaiser M, et al. Daily repetitive sensory stimulation of the paretic hand for the treatment of sensorimotor deficits in patients with subacute stroke: RESET, a randomized, sham-controlled trial. BMC neurology 2018; 18: 2. 2018/01/11. DOI: 10.1186/s12883-017-1006-z. 57. Kwong PW, Ng GY, Chung RC, et al. Transcutaneous electrical nerve stimulation improves walking capacity and reduces spasticity in stroke survivors: a systematic review and meta- analysis. Clinical rehabilitation 2017: 269215517745349. 2017/12/14. DOI: 10.1177/0269215517745349. 58. Wattchow KA, McDonnell MN and Hillier SL. Rehabilitation Interventions for Upper Limb Function in the First Four Weeks Following Stroke: A Systematic Review and Meta-Analysis of the Evidence. Arch Phys Med Rehabil 2017 2017/07/25. DOI: 10.1016/j.apmr.2017.06.014. 59. Hayward K, Barker R and Brauer S. Interventions to promote upper limb recovery in stroke survivors with severe paresis: a systematic review. Disability and rehabilitation 2010; 32: 1973-1986. 2010/10/23. DOI: 10.3109/09638288.2010.481027. 60. Kwakkel G, Lannin NA, Borschmann K, et al. Standardized measurement of sensorimotor recovery in stroke trials: Consensus-based core recommendations from the Stroke Recovery and Rehabilitation Roundtable. International journal of stroke : official journal of the International Stroke Society 2017; 12: 451-461. 2017/07/13. DOI: 10.1177/1747493017711813. 61. Stinear CM, Byblow WD, Ackerley SJ, et al. PREP2: A biomarker-based algorithm for predicting upper limb function after stroke. Ann Clin Transl Neurol 2017; 4: 811-820. 2017/11/22. DOI: 10.1002/acn3.488.

ESS in early rehabilitation post-stroke

Table 1: Description of the achieved ESS-intervention, usual rehabilitation (including arm training), and actual time points for post- intervention and follow-up assessments

Low-dose High-dose p-value group (n=49) group (n=53) ESS-intervention* Potential ESS-sessions, days, median (Q1―Q3) 13 (8―18) 13 (10―20) 0.45 Completed ESS-sessions, days, median (Q1―Q3) 11 (6―15) 11 (7―15) 0.80 Partially completed ESS-sessions, days, median (Q1―Q3) 1(0―1) 1(0―2) 0.15 Per-protocol ESS-interventions*, days, median (Q1―Q3) 9 (5―13) 10 (3―13) 0.28 ESS-interventions not given per protocol, days, median (Q1―Q3) 2 (1―4) 2 (1―4) 0.84 Total ESS-interventions (per protocol + not per protocol), days, median 12 (8―16) 13 (9―18) 0.33 (Q1―Q3) Arm training (including arm training following ESS) Arm training during the ESS-intervention period, hours, median (Q1―Q3) 4 (3―5.6) 5.3 (4.3―6.1) 0.02 Arm training after the ESS-intervention period, hours, median (Q1―Q3) 6.2 (0.9―19.8) 9 (2―18.8) 0.63 Total arm training during the trial, hours, median (Q1―Q3) 10.1 (4.9―25.1) 15.7 (6.6―24.4) 0.33 Usual rehabilitation (including arm training) OT/PT training during the ESS-intervention period, hours, median (Q1―Q3) 5.8 (4.6―9.5) 7.1 (5.3―9.5) 0.29 OT/PT training after the ESS-intervention period, hours, median (Q1―Q3) 24.9 (10.5―56.3) 27 (1.80―39.5) 0.75 Total OT/PT training during the trial, hours, median (Q1―Q3) 30.9 (16.0―65.0) 34.4 (16.4―47.7) 0.82 Other physical stroke rehabilitation during the trial, n (%) 10 (23.3) 12 (23.5) 1.00 Actual time points for post-intervention and follow-up assessments No. of days from stroke onset to post-intervention assessment, median 19 (13―26) 20 (15―26) 0.52 (Q1―Q3) No. of days from last ESS-session to post-intervention assessment, median 1 (1―4) 1 (1―3) 0.64 (Q1―Q3) No. of days from 6 months post-stroke to the scheduled date for follow-up 3.5 (0―14) 5 (1―16) 0.32 assessment, median (Q1―Q3) *ESS-intervention = ESS-session followed by minimum 15-minute arm training p≤0.01

ESS in early rehabilitation post-stroke

Table 2: Baseline demographic and clinical characteristics

Low-dose High-dose p- group (n=49) group (n=53) value Demographic characteristics Age, years, median (Q1–Q3) 71(64–80) 72 (64–79) 0.93 Sex, men, n (%) Men 25 (51) 28 (52.8) 1.00 Women 24 (49) 25 (47.2) Living arrangement, n (%) Living alone 28 (57.1) 34 (64.2) 0.54 Living with othersa 21(42.9) 19 (35.9) Actual profession/last profession before retirement, n (%) Non-manual workersb 29 (59.2) 33 (62.3) 0.65 Self-employedc 3 (6.1) 1 (1.9) Manual workersd 17 (34.7) 19 (35.9) Pre-stroke disability level (pre-stroke mRSe), mean 0.8 (1.1) 1 (1.3) 0.41 (SD), median (Q1–Q3) 0 (0–2) 0 (0–2) Stroke risk factors Smoking, n (%) Never smoker 8 (16.3) 15 (28.3) Former smoker 20 (40.8) 14 (26.4) 0.22 Current smoker 21 (42.9) 24 (45.3) High risk of disease due to alcohol consumptionf, 10 (20.4) 9 (17) 0.80 n (%) Increased to high risk of metabolic diseaseg, 22 (46.8) 25 (51) 0.69 n (%) Previous stroke, n (%) 12 (24.5) 10 (18.9) 0.63 Previous transient ischaemic attack, n (%) 1 (2) 1 (1.9) 1.00 Previous atrial fibrillation, n (%) 10 (20.4) 8 (15.1) 0.61 Previous myocardial infarction, n (%) 1 (2) 1 (1.9) 1.00 Previous angina pectoris, n (%) 0 (0) 0 (0) - Diabetes, n (%) 8 (16.3) 7 (13.2) 0.78 Psychiatric disorder, n (%) 3 (6.1) 4 (7.6) 1.00 Stroke in the family, n (%) 2 (4.1) 5 (9.4) 0.44 Heart failure, n (%) 8 (16.3) 3 (5.7) 0.11 Hypertension, n (%) 26 (53.1) 33 (62.3) 0.42 Myocardial infarction in the family, n (%) 2 (4.1) 3 (5.7) 1.00 Peripheral arterial disease, n (%) 1 (2) 1 (1.9) 1.00 Hyperlipidaemia, n (%) 9 (18.4) 15 (28.3) 0.25 Other diseases, n (%) 18 (36.7) 16 (30.2) 0.53 Pre-stroke physical activity level (PASEh), 76 (40–134) 88 (40–149) 0.58 median (Q1–Q3) Clinical characteristics Diagnosis, n (%) Ischaemic stroke (IS) 37 (75.5) 43 (81.13) 0.63 Haemorrhagic stroke (HS) 12 (24.5) 10 (18.87) TOAST classification of subtypes of ischemic stroke, n (%) Large-artery artherosclerosis 4 (10.8) 6 (14) Cardioembolism 9 (24.3) 8 (18.6) Small-artery occlusion 24 (64.9) 28 (65.1) 0.86 Stroke of undetermined etiology 0 (0) 1 (2.3) Stroke of other determined etiology 0 (0) 0 (0) Stroke severity (SSSi), n (%) ESS in early rehabilitation post-stroke

Mild to moderate stroke (SSS>25) 45 (91.8) 53 (100) 0.05 Major stroke (SSS≤25) 4 (8.2) 0 (0) Voluntary finger extension, yes, n (%) 37 (75.5) 42 (79.3) 0.81 Affected upper limb, right, n (%) 28 (57.1) 25 (47.2) 0.33 Dominant hand, right, n (%) 47 (95.9) 53 (100) 0.23 Affected dominant hand, n (%) 26(53.1) 25(47.2) 0.69 Sensory dysfunctions, (FMA-UE-H), yes, n (%) 26 (57.8) 28 (52.8) 0.69 Joint pain, (FMA-UE-J), yes, n (%) 17 (37) 20 (37.7) 1.00 Acute treatment, n (%) Thrombolysis 6 (12.3) 9 (17) Aspirin 26 (53.1) 34 (64.2) Thrombectomy 0 (0) 0 (0) 0.08 Other 5 (10.2) 0 (0) Not relevant (for HS) 12 (24.5) 10 (18.9) Medications at hospital admission, n (%) Aspirin 8 (16.33) 9 (17) 1.00 Persantin 2 (4.08) 0 (0) 0.23 Anticoagulation medications 8 (16.33) 6 (11.3) 0.57 Clopidogrel 6 (12.24) 9 (17) 0.58 High blood pressure medications 29 (59.18) 32 (60.4) 1.00 Lipid lowering medications 12 (24.49) 14 (26.4) 1.00 Other 36 (73.47) 34 (64.2) 0.39 Blood pressure (BP) (mmHg), median (Q1–Q3) Systolic BP (SBP) 148 (135–170) 153 (135–171.5) 0.56 Diastolic BP (DBP) 78 (69–87) 81.5 (72–89) 0.35 Pulse (/min), median (Q1–Q3) 71 (62–80) 76.5 (62–82.5) 0.41 Blood tests Total cholesterol ≥ 5.0 mmol/l 20 (46.5) 33 (67.4) 0.06 eGfR ≤ 60 ml/min 14 (30.4) 15 (28.3) 0.83 Glucose ≥ 7.7 mmol/l 10 (23.8) 6 (13.3) 0.27 aMarried, house share, sheltered housing, nursing home. bBusiness executives, engineers with university degrees, physicians, college teachers, secondary school teachers, office assistants, sales people cProfessionals with and without employees, entrepreneurs, farmers dAuto mechanics, metal workers, construction workers, factory workers, waiters, cleaning staff emodified Rankin Scale (0–5 points) (Sex=male AND alcohol consumption > 21 drinks/week) OR (Sex=female and alcohol consumption > 14 drinks/week) f(Sex=female AND 25≤BMI<30 AND waist circumference<88 cm) OR (Sex=male AND 25≤BMI<30 AND waist circumference<102 cm) OR (Sex=female AND 25≤BMI<30 AND waist circumference≥88 cm) OR (Sex=male AND 25≤BMI<30 AND waist circumference≥102 cm) OR (Sex=female AND BMI≥30) OR (Sex=male AND BMI≥30). Waist circumference measured in standing position gPhysical Activity Scale for Elderly hClassification of subtype of acute ischemic stroke iScandinavian Stroke Scale jFugl-Meyer Assessment-upper extremity section, subscale H (Sensation, 0-12) kFugl-Meyer Assessment-upper extremity section, subscale J (Joint pain during passive movement, 0−24) *p-value≤0.01 ESS in early rehabilitation post-stroke

Table 3: Between-group differences

Low-dose group High-dose group Effect size (L-H) Mean (SD) Mean (SD) Adjusted difference Median (Q1–Q3) Median (Q1–Q3) [95% CI], p-value Baseline Post Follow-up Baseline Post Follow-up Post Follow-up BBT 10.2 (12.8) 19.9 (16.2) 26.7 (16.7) 9.3 (11) 19.2 (15.2) 25.9 (15.7) -0.8 [-3.9; 2.3], -0.5 [-5.2;4.1], (blocks/min) 1 (0–17) 23 (3–32) 28 (14–39) 3 (0–16) 17 (6–28) 27 (16–38) p=0.62 p=0.82 FMA-UE-AD 33.2 (21) 46.7 (20.6) 51.2 (18.4) 33.6 (19.1) 44 (19.5) 50.7 (17.2) 1.7 [-3.2; 6.7], -0.9 [ -6.7; 5], (0−66 points) 37 (10–55) 56 (38–62) 58 (49.5–64) 37 (15– 50) 49.5 (36–59) 58 (46–63) p=0.49 p=0.77 PTT 7 (6.7) 5.1 (5.4) 4.6 (3.2) 5.6 (4.6) 4.2 (2.3) 3.7 (1.8) -0.5 [-2.1; 1.1], -0.3 [ -2.5; 1.9], (mA) 4.5 (3.5–7.4) 3.3 (2.8–5.0) 3.5 (2.9–4.3) 4.5 (3.5–6) 3.8 (2.8–5) 3.3 (2.6–4.3) p=0.53 p=0.80 Hand grip 12.0 (13) 16.5 (12.8) 18.4 (13.4) 11.0 (10) 14.9 (10.9) 18.1 (10.5) -0.1 [ -2.2; 2], -1.6 [-4.3; 1.2], strength (kg) 7.1 (0.7–12) 14.8 (5.1–25.4) 16.6 (6.5–28.7) 8.8 (2.6–16.1) 15.1 (6–22.1) 17.7 (9.9–24.4) p=0.92 p=0.27 Palmar pinch 1.9 (2.4) 2.9 (2.7) 3.3 (2.5) 1.4 (2.4) 2.3 (2.1) 3.1 (2.3) 0.0 [-0.5; 0.5], -0.3 [-1.1; 0.5], strength (kg) 0.5 (0–3.2) 2.3 (0–4.9) 3.5 (1.5–5.4) 0.7 (0–2.4) 2.4 (0–4.0) 3.1 (1.6–4.4) p=0.93 p=0.44 Key pinch 2.7 (3) 4.0 (3) 5.0 (3) 2.3 (3) 3.7 (2.6) 5 (2.9) -0.3[-1; 0.4], -0.1 [-1; 0.9], strength (kg) 2 (0–4.3) 4.0 (0.9–5.9) 4.5 (2.5–7.1) 1.3 (0–4.3) 3.7 (1.4–5.6) 4.5 (2.2–6.9) p=0.44 p=0.88 Tip pinch 1.4 (1.8) 2.4 (2.1) 2.7 (1.9) 1.3 (1.4) 1.8 (1.6) 2.3 (1.6) 0.3 [-0.1; 0.7], 0.2 [-0.3; 0.8], Strength (kg) 0.5 (0–2.6) 2.1 (0.5–4.1) 2.7 (1.1–4.5) 0.6 (0–2.5) 1.9 (0–2.7) 2.2 (1.0–3.4) p=0.12 p=0.38 NRS 0.6 (1.7) 0.2 (0.9) 0.5 (1.9) 0.3 (0.9) 0.3 (1.5) 0.3 (1) -0.3 [-0.9; 0.2], -0.1 [-0.7; 0.4], (0−10 points) 0 (0–0) 0 (0–0) 0 (0–0) 0 (0–0) 0 (0–0) 0 (0–0) p=0.20 p=0.62 mRS 4.5 (0) 3.2 (0.9) 2.4 (1) 4.5 (0) 3.1 (0.9) 2.5 (1) 0 [-0.3; 0.4], -0.2 [-0.6; 0.3], (0−5 points) 4.5 (4.5–4.5) 3 (2–4) 2 (2–3) 4.5 (4.5–4.5) 3 (3–4) 2 (2–4) p=0.82 p=0.49 Adjusted difference: Adjusted for sex and ability of voluntary finger extension at baseline; BBT: Box and Block Test; FMA-UE-AD: Upper extremity section of Fugl-Meyer Assessment, subscale A−D; H: High-dose group; L: Low-dose group; mRS: Modified Rankin Scale (0−5); NRS: Numerical Scale-11; Post: Post-intervention; PTT: Perceptual threshold of touch; *p≤0.01 (for differences in BBT: p≤0.05)

ESS in early rehabilitation post-stroke

Table 4: Within-group differences

Low-dose group High-dose group Adjusted difference Adjusted difference [95% CI], p-value [95% CI], p-value Baseline to Baseline to follow-up Baseline to Baseline to follow-up post-intervention post-intervention BBT 8.9 [ 6.8; 11] 16.0 [12.7; 19.1] 9.6 [ 7.4; 11.9] 16.5 [13.1; 19.9] (blocks/min) p<.0001* p<.0001* p<.0001* p<.0001* FMA-UE-AD 12.4 [ 8.3; 16.5] 16.6 [12.6; 20.6] 10.7 [8; 13.4], 17.5 [ 13.2; 21.8] (0−66 points) p<.0001* p<.0001* p<.0001* p<.0001* PTT -2 [-3.4; -0.6] -2.3 [-4.1; -0.6] -1.4 [-2.4; -0.6] -2.0 [-3.4; -0.7] (mA) p=0.003* p=0.008* p=0.001* p=0.002* Hand grip 3.7 [2.2; 5.3] 5.6 [3.9; 7.4] 3.8 [2.5; 5.2] 7.2 [5; 9.4] strength (kg) p<.0001* p<.0001* p<.0001* p<.0001* Palmar pinch 0.8 [0.5; 1.2] 1.3 [0.8; 1.8] 0.9 [0.5; 1.2] 1.6 [1; 2.2] strength (kg) p<.0001* p<.0001* p<.0001* p<.0001* Key pinch 1 [0.5; 1.5] 2.1 [1.5; 2.6] 1.3 [0.7; 1.8] 2.1 [1.4; 2.9] strength (kg) p<.0001* p<.0001* p<.0001* p<.0001* Tip pinch 0.8 [0.5; 1.1] 1.1 [0.7; 1.5] 0.5 [0.2; 0.7] 1 [0.4; 1.3] strength (kg) p<.0001* p<.0001* p<.0001* p<.0001* NRS -0.3 [-0.6; 0] -0.1 [-0.5; 0.3] 0 [ -0.4; 0.5] 0.1 [-0.3; 0.4] (0−10 points) p=0.076 p=0.723 p=0.850 p=0.728 mRS -1.3 [-1.6; -1.1] -2.1 [-2.4; -1.8] -1.4 [-1.6; -1.1] -1.9 [-2.2; -1.6] (0−5 points) p<.0001* p<.0001* p<.0001* p<.0001* Adjusted difference: Adjusted for sex and ability of voluntary finger extension at baseline; BBT: Box and Blocks Test; FMA AD=FMA-UE-AD: Upper extremity section of Fugl-Meyer Assessment, subscale A−D; NRS: Numerical Scale-11; PTT: Perceptual threshold of touch; mRS: Modified Rankin Scale; *p≤0.01 (for differences in BBT: p≤0.05); Differences ≥ MCID (if established) are in bold.

ESS in early rehabilitation post-stroke

Table 5: Additional results

Low-dose High-dose p-value group (n=49) group (n=53) Adverse reactions to ESS, n (%) 1 (2) 1 (1.9) 1.00 Length of hospital stay, days, median (Q1―Q3) 17 (12―23) 18 (15―22) 0.69 Recurrent stroke during participation in the study, n (%) 3 (6.1) 2 (3.8) 0.67 Complications during the hospital stay, n (%) Pneumonia 8 (16.3) 2 (3.8) 0.05 Urinary tract inflammation 14 (28.6) 14 (26.4) 0.83 Deep vein thrombosis 1 (2) 2 (3.8) 1.00 Pulmonary embolism 0 (0) 1 (1.9) 1.00 Fall 2 (4.1) 1 (1.9) 0.61 Discharge destination, n (%) Home 19 (38.8) 24 (45.3) Outpatient rehabilitation center in the community 1 (2) 1 (1.9) Inpatient rehabilitation center in the community 25 (51) 28 (52.8) 0.52 Nursing home, sheltered housing 1 (2) 0 (0) Hospital 1 (2) 0 (0) Dead 2 (4.1) 0 (0) Residence at follow-up, n (%)* Home 37 (78.7) 41 (80.4) Outpatient rehabilitation center in the community 0 (0) 0 (0) Inpatient rehabilitation center in the community 2 (4.3) 0 (0) 0.53 Nursing home, sheltered housing 4 (8.5) 6 (11.8) Hospital 0 (0) 1 (2) Dead 4 (8.5) 3 (5.9) p≤0.01

ESS in early rehabilitation post-stroke

SUPPLEMENTARY MATERIAL

A single ESS-session: the difference between high-dose and low-dose ESS

Paper III

Can clinical tests early post-stroke aid the prediction of arm functioning at 6 months?

Emma Ghaziani1, Christian Couppé 1,2, Volkert Siersma3, Hanne Christensen4, S. Peter

Magnusson1,2, Katharina S. Sunnerhagen5, Hanna C. Persson5, Margit Alt Murphy5

Author affiliations

1Department of Physical and Occupational Therapy, Bispebjerg and Frederiksberg Hospital, and

Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark

2Institute of Sports Medicine, Department of Orthopaedic Surgery M, Bispebjerg and Frederiksberg

Hospital, and Center for Healthy Aging, Faculty of Health and Medical Sciences, University of

Copenhagen, Copenhagen, Denmark

3The Research Unit for General Practice and Section of General Practice, Department of Public

Health, University of Copenhagen, Copenhagen, Denmark

4Department of Neurology, Bispebjerg and Frederiksberg Hospital, and Faculty of Health and

Medical Sciences, University of Copenhagen, Copenhagen, Denmark

5Department of Clinical Neuroscience and Rehabilitation, Institute of Neuroscience and Physiology,

Sahlgrenska Academy, University of Gothenburg, Sweden

Corresponding author

Emma Ghaziani

Bispebjerg and Frederiksberg Hospital, Department of Physical and Occupational Therapy, Nielsine

Nielsens Vej 10, 2400-Copenhagen NW, Denmark, [email protected]

ABSTRACT

Background: To be applicable in clinical practice, prognosis of arm recovery after stroke needs to be based on easy-to-perform, meaningful measures. Several clinical tests have been proposed for prognosis of arm functioning; further validation of their predictive value is needed.

Objective: To examine the individual predictive value of easy-to-perform clinical tests for early prognosis of arm functioning.

Methods: We performed a secondary analysis of merged data from two independent studies

(n=223). Following variables measured 3−7 days post-stroke using the Fugl-Meyer Assessment of

Upper Extremity (FMA-UE) were considered as potential predictors: shoulder abduction and elbow extension within synergies, forearm pronation/supination, wrist dorsiflexion, mass finger extension, grasping ability and the sensory subscale. Logistic regression was used for each predictor to calculate the odds ratio of two levels of arm functioning measured with the motor part of FMA-UE at 6 months post-stroke: FMA-UE-motor ≥32 and ≥58.

Results: Patients with partial shoulder abduction were at least 7.3 times more likely to achieve a

FMA-UE-motor≥32 at 6 months post-stroke. The probability of a FMA-UE-motor≥58 was at least

3.3−35.2 times higher in patients with partial/full distal movement (forearm pronation/supination, wrist dorsiflexion and grasping ability) compared with patients with absent movement. Patients with full elbow and finger extension were at least 36.8 and 18.3, respectively, times more likely to achieve a FMA-UE-motor≥58 than patients with no movement. Full sensory function had a significant, but more modest predictive value.

Conclusions: This study confirmed that sufficient sensory function and some proximal/distal arm movement early post-stroke predict a better arm functioning at 6 months in patients with moderately impaired arm function and mild-to-moderate stroke; partial/full distal movement was identified as predictor of a FMA-UE-motor≥58.

TRIAL REGISTRATION: ClinicalTrials.gov: NCT02250365 (ESS-trial), NCT01115348

(SALGOT-study)

INTRODUCTION

Stroke remains a major cause of disability worldwide (1), and arm paresis is one of the most prevalent consequences of stroke. Longitudinal cohort studies have shown that arm paresis is present in 48−77% of patients at stroke onset (2-4). Despite rehabilitation, only 12−34% of these will achieve a complete arm recovery at 6 months post-stroke (5, 6) which is the time point where the restoration of arm functioning seems to reach a plateau (7, 8). It has been demonstrated that an almost complete arm recovery is required for stroke patients to routinely involve their affected arm in the performance of activities of daily living (ADL) (9). It is therefore problematic that only a low percentage of these patients regain functional use of the affected arm. The presence of arm paresis impedes the patient’s independence in ADL (10) and affects the health-related quality of life detrimentally (11). Consequently, facilitating arm recovery is an essential issue in stroke rehabilitation, especially during the first 4−10 weeks when the restoration process is most pronounced (7, 8, 12-14). Physical and occupational therapists need to evaluate two essential aspects when planning the rehabilitation of the affected arm. Firstly, it is important to know the recovery potential for the arm, and the optimal proportion of restorative and compensatory therapeutic strategies to support this recovery in each individual patient. Secondly, therapists are expected to inform patients and their families about the long-term arm functioning level, allowing them to adjust to the new life situation and to plan for the future. Therefore, it is of relevance to make an accurate, early prediction of the long-term arm functioning from easily performed, clinical, bedside tests that are time-effective, suitable for most of the patients regardless their motor and

cognitive impairments, and do not rely on technologies that are inaccessible to the most hospital departments.

Several measures easily collected during clinical bedside examinations in the acute/early subacute stroke phase have been proposed as predictors for later levels of arm functioning. It has been shown that some degree of volitional finger extension measured on day 7 after stroke is an indicator of better arm function at 14 days, 1 month, 3 months and 6 months post- stroke (15); hand grip strength within the first week is associated with hand dexterity at 5 weeks post-stroke (16). Combinations of variables assessing proximal (i.e. shoulder abduction, elbow flexion) and distal (i.e. finger extension, hand grip strength, pinch grip) arm movement from day 3

(6) to 1 month (17-19) are likewise predictors of arm function and hand dexterity at 6 months (6,

17) and ≥1 year (18, 19) post-stroke. Furthermore, proximal arm movement alone measured at 1 month seems to predict arm function 3−4 months post-stroke (20). The presence of sensory dysfunction during the first month predict a poorer recovery of the affected arm’s function and dexterity at 6 months (17, 21) and ≥1 year after stroke (18).

In a systematic review with meta-analysis of predictors of arm recovery after stroke it was shown that there was strong evidence that higher arm function and functional capacity at baseline were associated with a better arm recovery (22). In line with recommendations from other reviews (23, 24), the authors emphasized the need for additional, high-quality cohort studies to validate these findings. The presence of motor and somatosensory evoked potentials was also a strong predictor of better arm recovery (22); though, these assessments are not widely applied in clinical practice, since they require special equipment and expertise. Consequently, the authors recommended the development and validation of clinical predictors which are more useful for clinical practice than neurophysiological measures.

The objective of this study was to examine the individual predictive value of easy-to- perform, clinical, bedside tests for early prognosis of arm functioning at 6 months post-stroke.

Specifically, we aimed to determine the value of proximal and distal arm movement, and sensory function assessed by the Fugl-Meyer Assessment of Upper Extremity (FMA-UE) (25) within 7 days after stroke for prognosis of a moderate and a mild motor impairment level at 6 months post-stroke.

We expected that our data would validate earlier proposed clinical tests such as shoulder abduction, finger extension, and sensory function for prediction of arm functioning after stroke.

This study is presented according to the guidelines for transparent reporting of multivariable prediction models for individual prognosis proposed by the TRIPOD-statement (26).

METHODS

Source of data

This study was a secondary analysis of merged data from two published, independent studies. The first study was a prospective longitudinal cohort study, the Stroke Arm Longitudinal

Study at the University of Gothenburg, Sweden (the SALGOT-study), that aimed to describe the recovery of arm functioning during the first 12 months after stroke (27). The second study was a randomized controlled trial conducted at the stroke rehabilitation unit of Bispebjerg and

Frederiksberg Hospital, Copenhagen, Denmark, and examined the effect of electrical somatosensory stimulation (ESS) delivered prior to task-oriented arm training during early inpatient rehabilitation on the recovery of arm functioning at 6 months post-stroke (the ESS-trial) (28). The

ESS-trial demonstrated no difference between the intervention and the control group (provisionally accepted for publication), making it possible for the present study to collapse these two trial groups into a cohort of stroke patients.

Trial settings and participants

In the SALGOT-study (27), all patients consecutively admitted at the largest of the three stroke units of the Sahlgrenska University Hospital, Gothenburg, from February 2009 to

December 2010 were screened using following inclusion criteria: a) first-ever stroke (ischaemic stroke, IS, or haemorrhagic stroke, HS); b) age ≥18 years; c) impaired arm function at day 3 (±1 day) after stroke (FMA-UE< 66); d) admission to the stroke unit ≤3 days after stroke onset; and e) residence in the Gothenburg urban area. Patients were excluded if one of the following criteria was present: a) injury/condition prior to the stroke that limited the use of the affected arm; b) severe, multiple impairments or diminished physical condition prior to stroke; c) short life expectancy; and d) not able to communicate in Swedish.

In the ESS-trial, all patients consecutively admitted to the stroke rehabilitation unit of

BBH from October 2014 to March 2017 (except a total of a 6-months period of holidays and recruitment/training of new trial staff) were assessed for eligibility according to following criteria: a) acute IS or HS; b) age ≥ 18 years; c) impaired arm function (FMA-UE< 66); d) residence in the hospital’s catchment area for stroke rehabilitation; e) no severe pre-stroke disability (modified

Rankin Scale =5); f) no contraindications to ESS (pacemaker, skin impairment); g) possibility for initiating the ESS-intervention within 7 days post-stroke due to medical or logistical issues; h) no cognitive dysfunctions or poor communication skills in Danish that impeded the ability to provide informed consent; i) complete recovery of the affected arm from a previous stroke; and j) no participation in other biomedical intervention trials within the last 3 months.

Outcome

The predicted outcome was two different levels of residual motor impairment assessed by the motor part of the FMA-UE (subscale A−D, 0−66 points) (FMA-UE-motor) (25, 29) at 6

months post-stroke: a) FMA-UE-motor ≥32 points, and b) FMA-UE-motor ≥58 points. It has been demonstrated that stroke patients with an arm impairment level <32 on FMA-UE-motor are unlikely to be able to perform basic ADL such as drinking from a glass with their affected arm (30, 31).

Furthermore, a minimum of 58 points on FMA-UE-motor has been suggested to represent the lower limit for a mild level of arm impairment (32), indicating a high probability that the patients routinely use the affected arm in performance of ADL due to an almost complete arm recovery (9).

The psychometric properties of the FMA-UE-motor have been extensively investigated. FMA-UE- motor has an excellent validity, and inter- and intra-observer reproducibility (33, 34). Among limitations, FMA-UE-motor seems to have a ceiling effect (34) as the most clinical scales, and it is considered too complex and time-consuming (30 minutes) for regularly use in clinical practice (35).

Potential predictors

We examined the predictive power of each of the following items from the FMA-UE- motor (25, 29) assessed during the first week post-stroke (baseline): a) shoulder abduction within synergies (subscale A.II), b) elbow extension within synergies (subscale A.II), c) forearm pronation/supination with 90-degree elbow flexion (subscale A.III), d) wrist stability at 15-degree dorsiflexion with 90-degree elbow flexion (subscale B), e) finger mass extension (subscale C), f) pincer grasp (subscale C), and g) cylinder grasp (subscale C). Each item is scored on a 3-level ordinal scale based on the clinician’s observation of the patient’s performance. Higher scores mean lesser impairment (0: absent, 1: partial, 2: full function). Furthermore, we considered the predictive value of the arm sensory function measured with the sensory part of the FMA-UE (subscale H,

0−12 points) (FMA-UE-sensory) (25, 29) and dichotomized as intact sensory function (FMA-UE- sensory =12) and sensory dysfunction (FMA-UE-sensory<12).

Sample size

The sample size for the present study was generated by merging data from the participants in the SALGOT-study (n=121) and the ESS-trial (n=102). For further details, see

Figure 1.

Missing data

Potential bias because of differential dropout and missing values was countered by weighting the remaining observations with the inverse of the estimated probability of this value being observed.

These estimated probabilities were for each observation estimated from a logistic regression model on the observation being missing or not including the following confounders: a) sample

(SALGOT/ESS), b) age, c) sex, d) living arrangement, e) previous stroke, f) stroke diagnosis, g) affected dominant hand, h) leg paresis, i) aphasia, j) stroke severity (Scandinavian Stroke Scale,

SSS) (36, 37), k) pre-stroke physical activity level (Saltin-Grimby Physical Activity Level Scale,

SGPALS-4) (38, 39), l) number of hospital days, and m) number of days from the stroke onset to the measurement of the potential predictors. The SALGOT-study assessed the stroke severity using the National Institutes of Health Stroke Scale (NIHSS-scale) (40), and the pre-stroke physical activity level with SGPALS-6 (41, 42). For the purpose of this study, the NIHSS-values were converted into SSS-values using the mathematical equation SSS=50.37–1.63xNIHSS (43); categories 1 and 2 of SGPALS-6 were merged into category 1 of GSPALS-4, and categories 5 and 6 of SGPALS-6 were merged into category 4 of SGPALS-4.

Statistical methods

Descriptive statistics were used for presenting demographics and clinical characteristics of the SALGOT-ESS-sample, and other confounders. We employed Wilcoxon Non-

Parametric Test and Fisher’s Exact Test for numerical and categorical variables, respectively, to determine whether the outcome groups (FMA-UE< 32/ ≥32, FMA-UE< 58/≥58) were statistically different with respect to these characteristics and other confounders.

For each potential predictor we used logistic regression to calculate the odds ratio of a favorable outcome (FMA-UE≥ 32/58) among patients with partial or full volitional motor function

(item score: 1 or 2), or full sensory function (FMA-UE-sensory=12) in comparison with the group of patients that presented no volitional motor function (item score: 0) or sensory dysfunction (FMA-

UE-sensory<12); the regressions were weighed as described in “Missing data”. Three logistic regression analyses were performed for each predictor: a) unadjusted; b) adjusted for: sample, sex, age and living arrangement, and c) adjusted for: sample, age, sex, living arrangement, previous stroke, stroke diagnosis, affected dominant hand, leg paresis, aphasia, SSS, SGPALS-4, number of hospital days, and number of days from the stroke onset to the measurement of the potential predictors. Analyses were performed with SAS version 9.4. The statistical significance was set to

1% to guard against false detection because of multiple comparisons.

RESULTS

Figure 1 details the inclusion process. The merged SALGOT-ESS-sample comprises

223 participants; and 121 from the SALGOT-study and 102 from the ESS-trial. The potential predictors and confounders were assessed at day 3 (median) (Table 1). In the SALGOT-study all baseline assessments were conducted at day 3 post-stroke; in the ESS-trial at day 5 (median). The outcome measure (FMA-UE-motor) was collected for 176/223 study participants at 6 months post- stroke.

Demographics and clinical characteristics of the SALGOT-ESS-sample at baseline, and other confounders are listed in Table 1. The median age was 71 years, and the proportion of

men was 54%. Half of the participants were living alone. The most prevalent risk factors for stroke were: a) hypertension: 54%, b) overweight: 42%, and c) physical inactivity before stroke onset:

25%. The majority of participants had a IS (82%) due to small-artery occlusion (43%), and presented a mild-to moderate stroke (87%). The arm function was moderately impaired (FMA-UE- motor=36, median); 63% had leg paresis and 22% aphasia. The median number of hospital days was 21. Furthermore, Table 1 presents the distribution of baseline demographics and clinical characteristics, and other confounders separately in two outcome groups (FMA-UE-motor< 32/

≥32, FMA-UE-motor< 58/≥58) at 6 months post-stroke. The groups with a less favourable outcome

(FMA-UE-motor<32/58) comprised a significantly higher percentage of participants with major strokes, leg paresis and more impaired arm motor function at baseline, and a longer hospital stay.

Table 2 and 3 show the main results of this study. Overall, the odds ratio (OR) for a favourable outcome (FMA-UE-motor≥ 32/58) was significantly higher among patients with partial/full motor function or full sensory function compared with patients with absent motor function or sensory dysfunction at baseline. Since the estimated effect sizes (OR a−c) are less accurate due to large 95% confidence intervals (95%CI), the lower limits of the 95%CI can be used as conservative estimates. Thus, when adjusting for all confounders (OR c), the probability of achieving a FMA-UE-motor ≥32 was at least 7.3-fold higher when the patients presented partial shoulder abduction within synergies compared with patients with no shoulder abduction. Whether forearm pronation/supination, wrist dorsiflexion, and grasping ability predict a FMA-UE-motor≥32 remains unanswered since the odds ratio could not be calculated due to the distribution of our data

(i.e. there were no patients with a score of 2 on these motor items at baseline and a FMA-UE-motor

< 32 at 6 months after stroke) (see Table 2). Moreover, the probability of achieving an almost complete arm recovery (FMA-UE-motor≥ 58) were at least 2.2−19-fold higher among patients showing partial elbow extension within synergies, forearm pronation/supination, wrist dorsiflexion,

and pincer and cylinder grasp without resistance, when OR was adjusted for all confounders (OR c).

In patients with full function on motor items, the probability of a FMA-UE-motor ≥58 at 6 months post-stroke was at least 5.7−36.8-fold higher compared with patients with no motor function at baseline. Patients with intact sensory function were at least 2.4-fold more likely to achieve a FMA-

UE-motor≥32 and at least 2.3-fold more likely to achieve a FMA-UE-motor≥58 at 6 months after stroke compared with patients with sensory dysfunction at baseline.

Figure 1: The flow of the participants through the study

ESS-SAMPLE SALGOT-SAMPLE

ASSESSED FOR ELIGIBILITY 1 ASSESSED FOR ELIGIBILITY: All patients consecutively admitted to the stroke All patients consecutively admitted to the stroke rehabilitation unit and unit and diagnosed with IS or HS from February 2009 to December 2010

diagnosed with IS or HS from October 2014 to March 2017 were screened for were screened for eligibility (n=763) newly developed arm paresis, except a total of 6 months (holidays, recruitment and training of new trial staff) (n=1,214)

EXCLUDED (n=641) • No arm impairment at day 1−2 post-stroke (n=335) ASSESSED FOR ELIGIBILITY 2 • Residence outside the hospital catchment area (n=56) Patients with a newly developed upper limb paresis (n=537) (Screening log • Prior arm impairment (n=58) for the initial 5 months is incomplete) • Severe multiple impairments (n=90) • Discharged<72h post-stroke (n=10) • Non-Swedish speakers (n=8) EXCLUDED (n=435) (Screening log for the initial 5 months is • Missed for screening (n=43) incomplete) • Decline to participate (n=36) • Not meeting remaining eligibility criteria (n=413) • Missed for inclusion (n=5) • Declined to participate (n=8) • Other reasons (n=14)

ASSESSEMENT OF CANDIDATE PREDICTORS AND CONFOUNDERS (3 DAYS POST-STROKE) (n=122) ASSE SSEMENT OF CANDIDATE PREDICTORS AND CONFOUNDERS • Not performed (n=2) (≤ 7 DAYS POST-STROKE) (n=102)

EXCLUDED: FMA-UE-motor=66 (3 days post-stroke) (n=1) SALGOT-ESS-SAMPLE (n=223)

• Not performed baseline assessment + refuse to participate (n=1) • Refused to participate (n=13) • Not followed-up (recurrent stroke, arm fracture) (n=11) • Dead (n=15) • Not performed baseline assessment + dead (n=1) • Not possible to contact (n=3) • Moved (n=3)

ASSESSMENT OF FMA-UE-motor (6 MONTHS POST-STROKE) (n=176)

Table 1: Demographics and clinical characteristics of the SALGOT-ESS-sample at baseline, and other confounders (n=223)

FMA-UE-motor at 6 months post-stroke < 32 ≥32 p-value < 58 ≥ 58 p-value Demographic characteristics Age, years, median (Q1−Q3) (min−max) 71 (63−80) 68 (62−78) 69 (62−80) 0.972 72 (62−80) 69 (62−78) 0.213 (26−95) (55−90) (34−92) (34−92) (38−92) Sex, n (%) Men 120 (54) 24 (65) 68 (52) 0.191 38 (51) 54 (57) 0.437 Women 103 (46) 13 (35) 64 (48) 37 (49) 40 (43) Living arrangement, n (%) Living alone 112 (50) 18 (49) 68 (52) 0.852 36 (48) 50 (53) 0.538 Living with others 111 (50) 19 (51) 64 (48) 39 (52) 44 (47) Risk factors for stroke Previous stroke, yes, n (%) 22 (10) 4 (11) 12 (9) 0.753 9 (12) 7 (7) 0.428 Previous transient ischaemic attack, yes, n (%) 2 (1) 0 (0) 2 (2) 1.000 1 (1) 1 (1) 1.000 Previous atrial fibrillation, yes, n (%) 45 (20) 7 (19) 22 (17) 0.805 14 (19) 15 (16) 0.684 Previous myocardial infarction, yes, n (%) 8 (4) 2 (5) 4 (3) 0.613 2 (3) 4 (4) 0.694 Previous angina pectoris, yes, n (%) 4 (2) 2 (5) 2 (2) 0.208 3 (4) 1 (1) 0.323 Diabetes, yes, n (%) 28 (13) 3 (8) 13 (10) 1.000 8 (11) 8 (9) 0.792 Psychiatric disorder, yes, n (%) 10 (4) 2 (5) 5 (4) 0.648 3 (4) 4 (4) 1.000 Heart failure, yes, n (%) 21 (9) 3 (8) 11 (8) 1.000 7 (9) 7 (7) 0.780 Hypertension, yes, n (%) 120 (54) 22 (59) 68 (52) 0.457 46 (61) 44 (47) 0.064 Peripheral arterial disease, yes, n (%) 3 (1) 1 (3) 2 (2) 0.525 2 (3) 1 (1) 0.585 Hyperlipidaemia, yes, n (%) 37 (17) 4 (11) 28 (21) 0.233 11 (15) 21 (22) 0.239 Other diseases, yes, n (%) 79 (35) 11 (30) 44 (33) 0.842 29 (39) 26 (28) 0.140 Overweight (BMI≥25), n (%) 93 (42) 15 (41) 66 (50) 0.354 36 (48) 45 (48) 1.000 Physically inactive pre-stroke (SGPALS=1), n (%) 53 (25) 5 (14) 31 (28) 0.461 20 (29) 16 (18) 0.256 Clinical characteristics Stroke diagnosis, n (%) Haemorrhagic stroke (HS) 41 (18) 7 (19) 29 (22) 0.821 19 (25) 17 (18) 0.263 Ischaemic stroke (IS) 182 (82) 30 (81) 103 (78) 56 (75) 77 (82) TOAST classification of subtypes of ischemic stroke, n (%) Large-artery artherosclerosis 28 (15) 4 (14) 18 (17) 0.565 9 (16) 13 (17) 0.714 Cardioembolism 51 (28) 11 (36) 23 (22) 16 (29) 18 (23) Small-artery occlusion 78 (43) 11 (36) 50 (49) 27 (49) 34 (44) Stroke of other determined etiology 10 (6) 2 (7) 6 (6) 2 (3) 6 (8) Stroke of undetermined etiology 15 (8) 2 (7) 6 (6) 2 (3) 6 (8) Acute treatment, yes, n (%) Thrombolysis 30 (13) 4 (11) 17 (13) 1.000 9 (12) 12 (13) 1.000 13

Thrombectomy 6 (3) 1 (3) 3 (3) 1.000 2 (3) 2 (2) 1.000 Stroke severity (SSS), n (%) Major stroke (SSS≤25) 29 (13) 10 (27) 7 (5) 0.000* 12 (16) 5 (5) 0.036 Mild-to-moderate stroke (SSS>25) 194 (87) 27 (73) 125 (95) 63 (84) 89 (95) FMA-UE-motor, median (Q1−Q3) (min−max) 36 (8−55) 4 (4−8) 47 (21−56) <.0001* 8 (4−18) 51 (41−58) <.0001* (0−65) (0−14) (0−65) (0−57) (4−65) Affected arm, right, n (%) 108 (48) 13 (35) 62 (47) 0.261 30 (40) 45 (48) 0.351 Dominant hand, right, n (%) 217 (97) 37 (100) 127 (96) 0.586 73 (97) 91 (97) 1.000 Affected dominant hand, yes, n (%) 106 (48) 13 (35) 61 (46) 0.263 28 (37) 46 (49) 0.160 Aphasia, yes, n (%) 49 (22) 10 (27) 24 (18) 0.250 19 (25) 15 (16) 0.176 Leg paresis, yes, n (%) 139 (63) 32 (86) 73 (56) 0.000* 61 (81) 44 (47) <.0001* Other confounders No. of hospital days, median (Q1−Q3) (min−max) 21 (13−34) 37 (22−44) 17 (12−30) <.0001* 31 (18−43) 16 (9−26) <.0001* (2−100) (10−100 (2−59) (8−100) (2−56) No. of days from the stroke onset to the measurement of the potential 3 (3−5) 3 (3−5) 3 (3−5) 0.228 3 (3−6) 3 (3−4) 0.117 predictors (baseline), median (Q1−Q3) (min−max) (0−7) (0−7) (2−7) (0−7) (2−7)

Abbreviations: BMI, Body Mass Index (44); SGPALS, Saltin-Grimby Physical Activity Level Scale; SSS, Scandinavian Stroke Scale (36); FMA-UE-motor, Fugl- Meyer Assessment of Upper Extremity (0-66 points) *: p≤0.01

Table 2: Odds ratio for a favorable outcome (FMA-UE-motor≥ 32) at 6 months

Candidate predictor, values FMA-UE-motor OR a 95%CI p OR b 95%CI p OR c 95%CI p (6 month) < 32 ≥32 Shoulder abduction, n 0: Absent=Reference 34 22 1: Partial 3 35 19.5 (5.3; 72.3) 0.000* 25.3 (5.7; 112.7) 0.000* 45.7 (7.3; 285.5) 0.000* 2: Full 0 74 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Elbow extension, n 0: Absent=Reference 35 22 1: Partial 2 31 32.9 (7.1; 152.1) 0.000* 43.8 (7.8; 246.7) 0.000* n.a. n.a. n.a. 2: Full 0 78 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Pronation/supination, n 0: Absent=Reference 37 30 1: Partial 0 40 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 2: Full 0 61 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Dorsiflexion, n 0: Absent=Reference 37 35 1: Partial 0 46 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 2: Full 0 50 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Finger extension 0: Absent=Reference 35 24 1: Partial 2 35 30.9 (6.7; 141. 9) 0.000* 36.1 (7.3; 178) 0.000* n.a. n.a. n.a. 2: Full 0 73 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Pincer grasp, n 0: Cannot grasp=Reference 37 48 1: Can grasp, but not hold against a tug 0 31 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 2: Can hold against a tug 0 52 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Cylinder grasp, 0: Cannot grasp=Reference 37 46 1: Can grasp, but not hold against a tug 0 28 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 2: Can hold against a tug 0 57 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. Sensory function, Sensory dysfunction=Reference 30 56 Full sensory function 7 73 6.4 (2.6; 15.8) 0.000* 8.8 (3; 26.4) 0.000* 9.8 (2.4; 39.4) 0.001*

Abbreviations: 95%CI: 95% confidence interval; FMA-UE-motor: Fugl-Meyer Assessment of Upper Extremity (0−66 points); n.a.: not applicable; OR a: odds ratio, unadjusted; OR b: odds ratio adjusted for sample, sex, age, living arrangement; OR c: odds ratio adjusted for sample, age, sex, living arrangement, previous stroke, 15

stroke diagnosis, affected dominant hand, leg paresis, aphasia, stroke severity, pre-stroke physical activity level, number of hospital days, and number of days from stroke onset to the assessment of candidate predictors. *: p≤0.01

Table 3: Odds ratio for a favorable outcome (FMA-UE-motor≥ 58) at 6 months post-stroke

Candidate predictor, values FMA-UE-motor OR a 95%CI p OR b 95%CI p OR c 95%CI p (6 months) <58 ≥58 Shoulder abduction, n 0: Absent=Reference 46 10 1: Partial 18 20 5.4 (2.1; 14) 0.000* 6.3 (2.2; 18.1) 0.000* 3.8 (1.2; 12.4) 0.026 2: Full 10 64 26.7 (10; 71.4) 0.000* 33.2 (11.8; 93.3) 0.000* 22 (5.7; 84.9) 0.000* Elbow extension, n 0: Absent=Reference 51 6 1: Partial 15 18 9.4 (3.1; 29) 0.000* 12.9 (3.8; 44.2) 0.000* 12.5 (2.2; 72.1) 0.004* 2: Full 8 70 77.9 (24.8; 244.2) 0.000* 121.1 (31.2; 470.6) 0.000* 567.5 (36.8; 8,762.8) 0.000* Pronation/supination, n 0: Absent=Reference 59 8 1: Partial 9 31 25 (8.6; 72.5) 0.000* 42.2 (11.3; 158) 0.000* 55.9 (8.7; 357.6) 0.000* 2: Full 6 53 72.6 (23.3; 226.2) 0.000* 171 (45.7; 639.6) 0.000* 367 (35.2; 3,826.5) 0.000* Dorsiflexion 0: Absent=Reference 63 9 1: Partial 8 38 32 (11.2; 91.3) 0.000* 41.1 (12.8; 132.4) 0.000* 118.6 (19; 740.4) 0.000* 2: Full 3 47 118.4 (28.3; 442.1) 0.000* 115.8 (25.8; 520.5) 0.000* 245.7 (24.5; 2,460.2) 0.000* Finger extension, n 0: Absent=Reference 53 6 1: Partial 15 22 12,1 (4; 36.3) 0.000* 19 (5.7; 63.1) 0.000* 7.9 (1.4; 44.4) 0.019 2: Full 7 66 74 (23; 238.8) 0.000* 97.9 (27.6; 346.5) 0.000* 113.2 (18.3; 702.5) 0.000* Pincer grasp, n 0: Cannot grasp=Reference 64 21 1: Can grasp, but not hold against a tug 7 24 10.4 (3.9; 28.1) 0.000* 12 (4.3; 33.6) 0.000* 14.8 (3.3; 67.2) 0.000* 2: Can hold against a tug 3 49 46 (12.9; 164.2) 0.000* 50 (13.7; 183.1) 0.000* 88.4 (9.5; 819.7) 0.000* Cylinder grasp, n 0: Cannot grasp=Reference 64 19 1: Can grasp, but not hold against a tug 5 23 17 (5.6; 51.5) 0.000* 21.6 (6.4; 73.3) 0.000* 37.6 (5; 285) 0.000* 2: Can hold against a tug 5 52 36.2 (12.5; 104.7) 0.000* 49.5 (13.4; 182.3) 0.000* 70 (10.6; 456.2) 0.000* Sensory function, n Sensory dysfunction=Reference 55 31 Full sensory function 17 63 7.1 (3.5; 14.4) 0.000* 6.8 (3.3; 14) 0.000* 6.1 (2.3; 16.7) 0.000*

Abbreviations: 95%CI: 95% confidence interval; FMA-UE-motor: Fugl-Meyer Assessment of Upper Extremity (0−66 points); OR a: odds ratio, unadjusted; OR b: odds ratio adjusted for sample, sex, age, living arrangement; OR c: odds ratio adjusted for sample, age, sex, living arrangement, previous stroke, stroke diagnosis, affected dominant hand, leg paresis, aphasia, stroke severity, pre-stroke physical activity level, number of hospital days, and number of days from stroke onset to the assessment of candidate predictors. *: p≤0.01

DISCUSSION

This study evaluated the individual value of clinical tests that are easy-to-perform early after stroke for prediction of arm functioning at 6 months post-stroke. The data show the presence of partial shoulder abduction within synergies measured 3−7 days after stroke was the strongest predictor (OR c=7.3) of the ability of involving the affected arm in performance of at least basic ADL at 6 months post-stroke. Furthermore, full elbow extension within synergies (OR c=36.8), full mass finger extension (OR c=18.3), partial/full wrist dorsiflexion (OR c=19/24.5) and partial/full forearm pronation/supination (OR c=8.7/35.2) were the strongest predictors of an almost complete arm recovery 6 months post-stroke.

The data suggest that partial/full proximal arm function (i.e. shoulder abduction and elbow extension within synergies), partial/full distal arm function (i.e. forearm pronation/supination, wrist dorsiflexion, pincer and cylinder grasp, mass finger extension), and full sensory function early after stroke are independently associated with a better arm functioning

(FMA-UE-motor≥ 32/58) at 6 months post-stroke. These results are in line with findings from previous studies showing that proximal arm control (6, 17-20), distal arm control (16-19), and sufficiently sensory function (17, 18, 21) measured during the first month predict a better hand dexterity at 3 to ≥12 months after stroke. To the best of our knowledge, this is the first investigation that aimed to specifically identify clinical tests that predict an almost complete arm recovery after stroke (FMA-UE-motor≥58).

Interestingly, the presence of shoulder abduction within synergies and finger mass extension during the first week after stroke was not significantly associated with an almost complete arm recovery at 6 months post-stroke in this study, unless the participants were able to perform these movements fully (item score=2).

The advantages of using the FMA-UE-motor as outcome allows predicting improvements predominantly due to pure neurologic recovery rather than functional recovery, which includes the interference of compensatory strategies (e.g. forward-bending of the upper body to compensate for impaired elbow extension in reaching tasks). Functional recovery might, however, be meaningful for patients and health care professionals since it reflects the patients’ ability to perform daily tasks relevant for their real-life. Thus, dichotomizing the FMA-UE-motor according to the proposed cut-off levels of 32 and 58 points, indicating the ability of performing at least basic ADL (30, 31) and the ability of routinely using the affected arm in ADL, respectively

(32), provides an opportunity to relate the outcome to functional ability, which is a strength of this study. Furthermore, it allows us to make inferences about the functional ability of the affected arm at 6 months post-stroke based on impairments assessed during the first week after stroke. The large sample size (n=223) achieved by merging two independent cohorts of stroke patients from two hospitals in different Scandinavian countries is another strength of this study, which improves the generalizability of the results. Finally, both the predictors and the predicted outcome are clinically relevant, easy-to-perform, and suitable for the majority of the stroke patients, which makes our study valuable for clinical practice.

In the present sample, the distribution of study participants with partial/full motor function at baseline was, with very few exceptions, 0 for an arm functioning of FMA-UE-motor<32 at 6 months post-stroke (see Table 2). Consequently, two limitations arised. Firstly, it was not possible to calculate the OR for a FMA-UE-motor≥32 for the majority of the motor items.

Secondly, the subgroup of patients with absent motor function (a score of 0 on items of FMA-UE- motor at baseline) and sensory dysfunction (a score <12 on FMA-UE-sensory at baseline) was selected as reference groups in the logistic regression analyses. It would have been more valuable for the clinicians to assess the predictive value of the selected clinical tests in the subgroup of

patients with absent/partial motor function and sensory dysfunction in comparison with a reference group consisting of patients with full motor/sensory function. However, this type of data distribution is not unique to the SALGOT- ESS-sample, but rather common in acute stroke trials.

CONCLUSION

The present study confirms previous findings that sufficient sensory function and some proximal/distal arm movement during the first week post-stroke are associated with a better arm functioning at 6 months post-stroke in patients with moderately impaired arm function and mild-to- moderate stroke. We further extend those finding by identifying individual clinical tests that specifically predict an almost complete arm recovery; we demonstrated that the presence of more demanding distal arm movements (i.e. partial/full forearm pronation/supination wrist dorsiflexion, grasping ability and full finger extension) during the first week after stroke was a predictor of an almost complete recovery 6 months post-stroke in patients with same characteristics. Further studies are needed to refine clinically accessible prediction models for patients who do not show any initial volitional movement early after stroke.

OTHER INFORMATION

Conflicting interests

The authors declare no conflict of interest.

Availability of data

Interested researchers may submit requests for the SALGOT-ESS-data set to EG and MAM

([email protected]) after obtaining permission from the Danish and Swedish

Committees on Health Research Ethics.

Funding

This work is based on merged data from two independent studies. The ESS-trial was supported by: a) the Capital Region of Denmark, Foundation for Health Research; b) Bevica Fonden; c) Lundbeck

Foundation [FP 68/2013], d) the Danish Association of Occupational Therapists [FF 1 14-3]; e)

Direktør Jacob Madsen & hustru Olga Madsen’s fond [5507], and f) the Department of Physical and

Occupational Therapy, Bispebjerg Hospital. The SALGOT-study received support from: a) the

Swedish Brain Foundation, b) the Swedish Heart and Lung Foundation, c) the local R&D Board for

Gothenburg and Södra Bohuslän, d) Promobilia, e) the Swedish National Stroke Association, and f) the Swedish Research Council (VR 2011-2718). The funding sources had no influence on design, data collection, analysis, interpretation or reporting of the results of the present study.

REFERENCES

1. Feigin VL, Krishnamurthi RV, Parmar P, Norrving B, Mensah GA, Bennett DA, et al. Update on the Global Burden of Ischemic and Hemorrhagic Stroke in 1990-2013: The GBD 2013 Study. Neuroepidemiology. 2015;45(3):161-76. 2. Lawrence ES, Coshall C, Dundas R, Stewart J, Rudd AG, Howard R, et al. Estimates of the prevalence of acute stroke impairments and disability in a multiethnic population. Stroke. 2001;32(6):1279- 84. 3. Nakayama H, Jorgensen HS, Raaschou HO, Olsen TS. Recovery of upper extremity function in stroke patients: the Copenhagen Stroke Study. Archives of physical medicine and rehabilitation. 1994;75(4):394-8. 4. Persson HC, Parziali M, Danielsson A, Sunnerhagen KS. Outcome and upper extremity function within 72 hours after first occasion of stroke in an unselected population at a stroke unit. A part of the SALGOT study. BMC neurology. 2012;12:162. 5. Kwakkel G, Kollen BJ, van der Grond J, Prevo AJ. Probability of regaining dexterity in the flaccid upper limb: impact of severity of paresis and time since onset in acute stroke. Stroke. 2003;34(9):2181-6. 6. Nijland RH, van Wegen EE, Harmeling-van der Wel BC, Kwakkel G. Presence of finger extension and shoulder abduction within 72 hours after stroke predicts functional recovery: early prediction of functional outcome after stroke: the EPOS cohort study. Stroke. 2010;41(4):745-50. 7. Kwakkel G, Kollen BJ. Predicting activities after stroke: what is clinically relevant? Int J Stroke. 2013;8(1):25-32. 8. Verheyden G, Nieuwboer A, De Wit L, Thijs V, Dobbelaere J, Devos H, et al. Time course of trunk, arm, leg, and functional recovery after ischemic stroke. Neurorehabilitation and neural repair. 2008;22(2):173-9.

9. Fleming MK, Newham DJ, Roberts-Lewis SF, Sorinola IO. Self-perceived utilization of the paretic arm in chronic stroke requires high upper limb functional ability. Archives of physical medicine and rehabilitation. 2014;95(5):918-24. 10. Sveen U, Bautz-Holter E, Sodring KM, Wyller TB, Laake K. Association between impairments, self-care ability and social activities 1 year after stroke. Disability and rehabilitation. 1999;21(8):372-7. 11. Chen CM, Tsai CC, Chung CY, Chen CL, Wu KP, Chen HC. Potential predictors for health- related quality of life in stroke patients undergoing inpatient rehabilitation. Health Qual Life Outcomes. 2015;13:118. 12. Kwakkel G. Impact of intensity of practice after stroke: issues for consideration. Disability and rehabilitation. 2006;28(13-14):823-30. 13. Lee KB, Lim SH, Kim KH, Kim KJ, Kim YR, Chang WN, et al. Six-month functional recovery of stroke patients: a multi-time-point study. International journal of rehabilitation research Internationale Zeitschrift fur Rehabilitationsforschung Revue internationale de recherches de readaptation. 2015;38(2):173-80. 14. van Kordelaar J, van Wegen E, Kwakkel G. Impact of time on quality of motor control of the paretic upper limb after stroke. Archives of physical medicine and rehabilitation. 2014;95(2):338-44. 15. Smania N, Paolucci S, Tinazzi M, Borghero A, Manganotti P, Fiaschi A, et al. Active finger extension: a simple movement predicting recovery of arm function in patients with acute stroke. Stroke. 2007;38(3):1088-90. 16. Higgins J, Mayo NE, Desrosiers J, Salbach NM, Ahmed S. Upper-limb function and recovery in the acute phase poststroke. Journal of rehabilitation research and development. 2005;42(1):65-76. 17. Au-Yeung SS, Hui-Chan CW. Predicting recovery of dextrous hand function in acute stroke. Disability and rehabilitation. 2009;31(5):394-401. 18. Kong KH, Chua KS, Lee J. Recovery of upper limb dexterity in patients more than 1 year after stroke: Frequency, clinical correlates and predictors. NeuroRehabilitation. 2011;28(2):105-11. 19. Kong KH, Lee J. Temporal recovery and predictors of upper limb dexterity in the first year of stroke: a prospective study of patients admitted to a rehabilitation centre. NeuroRehabilitation. 2013;32(2):345-50. 20. Houwink A, Nijland RH, Geurts AC, Kwakkel G. Functional recovery of the paretic upper limb after stroke: who regains hand capacity? Archives of physical medicine and rehabilitation. 2013;94(5):839- 44. 21. Winters C, Kwakkel G, Nijland R, van Wegen E. When Does Return of Voluntary Finger Extension Occur Post-Stroke? A Prospective Cohort Study. PloS one. 2016;11(8):e0160528. 22. Coupar F, Pollock A, Rowe P, Weir C, Langhorne P. Predictors of upper limb recovery after stroke: a systematic review and meta-analysis. Clin Rehabil. 2012;26(4):291-313. 23. Kwah LK, Herbert RD. Prediction of Walking and Arm Recovery after Stroke: A Critical Review. Brain Sci. 2016;6(4). 24. Harvey RL. Predictors of Functional Outcome Following Stroke. Phys Med Rehabil Clin N Am. 2015;26(4):583-98. 25. Fugl-Meyer AR, Jaasko L, Leyman I, Olsson S, Steglind S. The post-stroke hemiplegic patient. 1. a method for evaluation of physical performance. Scand J Rehabil Med. 1975;7(1):13-31. 26. Collins GS, Reitsma JB, Altman DG, Moons KG. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD statement. BMJ. 2015;350:g7594. 27. Alt Murphy M, Persson HC, Danielsson A, Broeren J, Lundgren-Nilsson A, Sunnerhagen KS. SALGOT--Stroke Arm Longitudinal study at the University of Gothenburg, prospective cohort study protocol. BMC neurology. 2011;11:56. 28. Ghaziani E, Couppe C, Henkel C, Siersma V, Sondergaard M, Christensen H, et al. Electrical somatosensory stimulation followed by motor training of the paretic upper limb in acute stroke: study protocol for a randomized controlled trial. Trials. 2017;18(1):84.

29. Fugl-Meyer AR JL, Leyman I, Olsson S, Steglind S. Fugl-Meyer Assessment - Upper Extremity (FMA-UE) University of Gotheburg. Institute of Neuroscience and Physiology1975 [updated 11 November 2015. Available from: https://neurophys.gu.se/english/departments/clinical_neuroscience_and_rehabilitation/rehabilitation- medicine/fugl-meyer. 30. Alt Murphy M, Willen C, Sunnerhagen KS. Kinematic variables quantifying upper-extremity performance after stroke during reaching and drinking from a glass. Neurorehabil Neural Repair. 2011;25(1):71-80. 31. Persson HC, Alt Murphy M, Danielsson A, Lundgren-Nilsson A, Sunnerhagen KS. A cohort study investigating a simple, early assessment to predict upper extremity function after stroke - a part of the SALGOT study. BMC neurology. 2015;15:92. 32. Pang MY, Harris JE, Eng JJ. A community-based upper-extremity group exercise program improves motor function and performance of functional activities in chronic stroke: a randomized controlled trial. Archives of physical medicine and rehabilitation. 2006;87(1):1-9. 33. Alt Murphy M, Resteghini C, Feys P, Lamers I. An overview of systematic reviews on upper extremity outcome measures after stroke. BMC neurology. 2015;15:29. 34. Gladstone DJ, Danells CJ, Black SE. The fugl-meyer assessment of motor recovery after stroke: a critical review of its measurement properties. Neurorehabil Neural Repair. 2002;16(3):232-40. 35. Barak S, Duncan PW. Issues in selecting outcome measures to assess functional recovery after stroke. NeuroRx. 2006;3(4):505-24. 36. Multicenter trial of hemodilution in ischemic stroke--background and study protocol. Scandinavian Stroke Study Group. Stroke. 1985;16(5):885-90. 37. Boysen G, Christensen H. Stroke severity determines body temperature in acute stroke. Stroke. 2001;32(2):413-7. 38. Grimby G, Borjesson M, Jonsdottir IH, Schnohr P, Thelle DS, Saltin B. The "Saltin-Grimby Physical Activity Level Scale" and its application to health research. Scand J Med Sci Sports. 2015;25 Suppl 4:119-25. 39. Saltin B, Grimby G. Physiological analysis of middle-aged and old former athletes. Comparison with still active athletes of the same ages. Circulation. 1968;38(6):1104-15. 40. Brott T, Adams HP, Jr., Olinger CP, Marler JR, Barsan WG, Biller J, et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke. 1989;20(7):864-70. 41. Grimby G. Physical activity and muscle training in the elderly. Acta Med Scand Suppl. 1986;711:233-7. 42. Mattiasson-Nilo I, Sonn U, Johannesson K, Gosman-Hedstrom G, Persson GB, Grimby G. Domestic activities and walking in the elderly: evaluation from a 30-hour heart rate recording. Aging (Milano). 1990;2(2):191-8. 43. Gray LJ, Ali M, Lyden PD, Bath PM, Virtual International Stroke Trials Archive C. Interconversion of the National Institutes of Health Stroke Scale and Scandinavian Stroke Scale in acute stroke. J Stroke Cerebrovasc Dis. 2009;18(6):466-8. 44. WHO. World Health Organization. BMI clasification 2006 [Available from: http://apps.who.int/bmi/index.jsp?introPage=intro_3.html.